<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by RAHULA RAJ on Medium]]></title>
        <description><![CDATA[Stories by RAHULA RAJ on Medium]]></description>
        <link>https://medium.com/@rahular2020?source=rss-8ed1df361839------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*6hvOy3Rj0UIQgP2bEbZ5sw.jpeg</url>
            <title>Stories by RAHULA RAJ on Medium</title>
            <link>https://medium.com/@rahular2020?source=rss-8ed1df361839------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 13:18:37 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@rahular2020/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[We Analyzed 207,297 AI Citations for Shopping Queries. Reddit Was Almost Nowhere.]]></title>
            <link>https://medium.com/@rahular2020/we-analyzed-207-297-ai-citations-for-shopping-queries-reddit-was-almost-nowhere-99743119de4b?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/99743119de4b</guid>
            <category><![CDATA[ai-visibility]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[shopify]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[seo]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Sun, 09 Aug 2026 06:13:44 GMT</pubDate>
            <atom:updated>2026-08-09T06:13:44.457Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*AfD2tO9KZsZnGiVk.png" /></figure><p><a href="https://apps.shopify.com/foundgpt-llms-txt">FoundGPT</a> analyzed 207,297 AI citations from 107 Shopify stores. Reddit was 0.5% of shopping citations — and ChatGPT cited it zero times. What AI actually cites when shoppers ask what to buy: product pages, not forums.</p><p><em>The most-repeated advice in AI search optimization is “get on Reddit.” Our data from 107 Shopify stores says that for shopping queries, that advice is wrong on the engine that matters most.</em></p><p>If you’ve read anything about Generative Engine Optimization (GEO) this year, you’ve seen the Reddit claim: Reddit is the single most-cited source in AI search, cited many times more than typical websites, so brands should invest in community presence. Industry studies back this up — for SaaS and technology queries.</p><p>Ecommerce is not SaaS. FoundGPT tracks what AI assistants actually answer when real shoppers ask real buying questions — <em>“best commuter e-bike under $2,000”</em>, <em>“non-toxic play mats for toddlers”</em> — across the engines shoppers actually use. Every answer’s sources get recorded and categorized.</p><p>This is what 207,297 citations say about where AI engines send shoppers. Spoiler: it isn’t Reddit.</p><h3>Key findings (TL;DR)</h3><ul><li><strong>Reddit was cited 1,032 times out of 207,297 citations — 0.5%.</strong> For shopping queries, Reddit is a rounding error, not a pillar.</li><li><strong>ChatGPT cited Reddit exactly zero times</strong> in our corpus. Not “rarely” — zero, across 58,846 ChatGPT citations and 3,601 shopping prompts. If your buyers use ChatGPT, a Reddit-first GEO strategy earns you nothing.</li><li><strong>Retailer and brand pages dominate every engine.</strong> On ChatGPT, product and store pages account for 67.5% of all citations (52.8% retailer + 14.7% brand sites). AI answers shopping questions with shopping pages.</li><li><strong>Where Reddit does appear, it’s Google AI Mode and Perplexity</strong> — 578 and 453 citations respectively. If your niche skews toward those engines, community matters. Otherwise it doesn’t.</li><li><strong>“Community” in shopping means YouTube, not forums.</strong> ChatGPT’s community citations are YouTube reviews, Goodreads, Pinterest and Lyst. YouTube alone out-cites Reddit two-to-one corpus-wide (2,189 vs 1,032).</li><li><strong>Independent Shopify stores can out-cite Amazon.</strong> Within the niche buyer prompts we track, two independent stores in our fleet each earned over 240 ChatGPT citations — amazon.com earned 22 in the same prompts. In its own niche, a well-structured independent store beats the giants by 10x. The field is open.</li></ul><h3>Scope of study</h3><p>The dataset: <strong>207,297 citation instances</strong> collected between <strong>April 19 and August 9, 2026</strong>, from <strong>24,216 visibility checks</strong> run for <strong>107 Shopify stores</strong> across <strong>3,601 distinct buyer-decision prompts</strong>, spanning <strong>17,695 unique cited domains</strong> on five engines: ChatGPT, Google Gemini, Google AI Mode, Perplexity, and Claude.</p><p>Three honesty notes, because a study you can’t trust is worthless:</p><ol><li><strong>This is descriptive, not causal.</strong> We report what engines cited, not why any single page was chosen.</li><li><strong>The prompt mix reflects our merchants</strong> — independent ecommerce niches (cycling, books, fashion, outdoor gear, wellness), weighted toward buyer-decision questions. A corpus built on electronics price-comparison prompts might look different.</li><li><strong>Claude is under-sampled in our corpus</strong> (158 citations — too few to report). For what it’s worth, OtterlyAI’s dedicated Claude study found Reddit at exactly 0% there too, on a 379,321-citation dataset.</li></ol><h3>Finding 1: AI answers shopping questions with shopping pages</h3><p>Share of citations by source category, per engine:</p><p><em>(Gemini’s high unclassified share is structural: Gemini masks 99.7% of its citation URLs behind redirect links — a finding big enough to get its own article.)</em></p><p>The pattern is unambiguous. When someone asks an AI assistant what to buy, the assistant reaches for <strong>pages where the product lives</strong>: retailer listings, marketplace pages, and the stores’ own sites. Editorial and community content fills in around the edges — it doesn’t lead.</p><p>This mirrors what OtterlyAI found for Claude in SaaS (64% brand domains) — but in shopping the effect extends to every engine we track.</p><h3>Finding 2: The Reddit number, engine by engine</h3><p>Reddit citations across the full corpus:</p><p>Sit with the ChatGPT row for a second. ChatGPT is the assistant your shoppers most likely use — it’s where most of the AI-referred sessions we measure on live storefronts come from. Across four months of shopping prompts, it cited Reddit <strong>zero times</strong>.</p><p>The “Reddit dominates AI search” finding is real — in SaaS and tech verticals, on specific engines. It just doesn’t transfer to ecommerce, and it especially doesn’t transfer to ChatGPT answering shopping questions. Porting that playbook to your store means investing in the one channel our data says pays nothing on the engine that matters most.</p><h3>Finding 3: Where Reddit does matter — and which communities</h3><p>On Google AI Mode and Perplexity, community content holds a real 16–19% share, and Reddit is a meaningful part of it. Which subreddits got cited?</p><p>(The subreddit mix partly reflects our merchants’ niches — that’s the point: <em>your</em> niche’s subreddits are what matter, not Reddit in general.)</p><p>Notice what these have in common: they’re <strong>hyper-specific enthusiast communities with substantive threads</strong> — “what gravel bike for a 100km commute” — not brand-promotion spaces. If you sell into a niche with a genuinely active enthusiast subreddit AND your buyers use Perplexity or Google’s AI results, authentic community presence can earn citations. That’s the narrow, honest version of the Reddit advice.</p><h3>Finding 4: “Community” in shopping means YouTube</h3><p>Top community sources cited (excluding Gemini’s masked URLs):</p><ul><li><strong>YouTube</strong> — #1 community source on <em>every</em> engine (784 on AI Mode, 600 on Perplexity, 324 on ChatGPT)</li><li><strong>Reddit</strong> — #2 on AI Mode and Perplexity, absent on ChatGPT</li><li><strong>Goodreads, Pinterest, Lyst, Instagram, Facebook</strong> — the rest of ChatGPT’s community mix</li></ul><p>Corpus-wide, YouTube (2,189) out-cites Reddit (1,032) two-to-one. If you’re going to invest anywhere off-site for AI visibility in ecommerce, a product review or comparison video is better-supported by the data than a Reddit thread — and it works on ChatGPT, where Reddit doesn’t.</p><h3>What this means for your store</h3><p><strong>1. Your product pages are the battlefield.</strong> Two-thirds of what ChatGPT cites for shopping queries is retailer and brand pages. A product page with a clear title, complete structured data, real descriptions, and visible pricing/availability is your single highest-leverage AI-visibility asset. Not your social presence. Not a forum campaign.</p><p><strong>2. You don’t need to be Amazon.</strong> Citations spread across 17,695 domains, and inside their own niches, independent Shopify stores in our fleet out-cite amazon.com by 10x on ChatGPT. AI engines reward the <em>most useful page for the specific question</em>, not the biggest domain.</p><p><strong>3. Right-size the Reddit effort.</strong> Selling in a niche with an active enthusiast subreddit, and your buyers use Perplexity or Google AI results? Genuine participation can earn citations. Everyone else: the hours are better spent on product pages and a YouTube review.</p><p><strong>4. Match the engine to your buyer.</strong> The engines behave differently enough that “AI search” isn’t one channel — ChatGPT is brand-and-retail, Perplexity and AI Mode lean community, Gemini hides its sources. Optimizing blind is guessing.</p><h3>Check where your store stands</h3><p>FoundGPT’s free scanner checks whether your store’s pages give AI engines what they cite — structured data, extractable content, the fundamentals this study shows actually get referenced: <a href="https://foundgpt.app/scan"><strong>foundgpt.app/scan</strong></a></p><p>Shopify merchants can track their own mentions and citations across all five engines — and see the actual sessions and orders AI sends them — with the <a href="https://apps.shopify.com/foundgpt-llms-txt">FoundGPT app</a>.</p><p><em>Methodology: citations were collected via FoundGPT’s visibility checks — live prompts run against each engine’s consumer surface, with every cited URL recorded and categorized (retailer, brand site, blog, community, news, review site). Counts are citation instances, not unique pages. We report counts and shares, never causal claims. Running your own GEO experiment on this? We’d love to compare data</em></p><h3>About the author</h3><p><strong>Rahul</strong> — Founder, FoundGPT</p><p>Rahul built FoundGPT and ran the analysis behind this study. Across 111 Shopify stores he has run 23,000+ real AI-search checks on ChatGPT, Gemini, Google AI Mode and Perplexity, and applied ~99,000 structured-data fixes.</p><p>→ <a href="https://foundgpt.app/about">About FoundGPT</a></p><p><em>Originally published at </em><a href="https://foundgpt.app/blog/reddit-ai-search-shopping-study"><em>https://foundgpt.app</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=99743119de4b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Is DeepSeek Better Than ChatGPT , Claude, Mistral, and Gemini for Pair Trading ?]]></title>
            <link>https://levelup.gitconnected.com/is-deepseek-better-than-chatgpt-claude-mistral-and-gemini-for-pair-trading-3de9b3d068d0?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/3de9b3d068d0</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[deepseek]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[trading]]></category>
            <category><![CDATA[finance]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Fri, 31 Jan 2025 16:35:40 GMT</pubDate>
            <atom:updated>2025-01-31T16:35:40.861Z</atom:updated>
            <content:encoded><![CDATA[<h3>Is DeepSeek Better Than ChatGPT , Claude, Mistral, and Gemini for Pair Trading</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d_Gqaw9iTwx7_QW2FDUfSQ.png" /><figcaption>Dalle</figcaption></figure><p>With Gen AI coming of age there are plethora of opportunities across sectors. One such sector is finance while financial statement is the most discussed. There is an opportunity of using Gen AI in trading as well.</p><p>In one of the courses that I am doing we were introduced to the concept of using Gen AI in Algorithm trading.In this article we will have a look on how the current models can be used and the limitations.</p><p>The current model be it ChatGPT, Gemini, Claude, Mistral are trained on different datasets which means same input would give a different output in each of these models. ChatGPT was trained on documents and research which makes it good for Factual analysis. Gemini is trained on used query and comments which makes it good for opinion based analysis.</p><p>Let’s deep dive on recommendation and create a algorithm based on recommendation. We are going to use pair wise trading which is built on the hypothesis that strongly correlated stock have similar return and any divergence would be corrected in long run.</p><h3>Prompt:</h3><p>suggest stock to do pair trading for Indian NSE from 2014- 2023 which have high correlation</p><h3>GenAI Models</h3><h4>DeepSeek</h4><p>Pair trading is a market-neutral strategy that involves identifying two stocks with a high historical correlation and trading them when the correlation deviates from its historical norm. To suggest pairs for pair trading on the Indian NSE from 2014 to 2023, we need to analyze historical data for stocks that have shown a high correlation over this period. Below are some steps and potential pairs based on historical trends and sectors:</p><h3>Steps to Identify Pair Trading Stocks:</h3><ol><li><strong>Sector Analysis</strong>: Stocks within the same sector often exhibit high correlation due to similar business fundamentals and market conditions.</li><li><strong>Historical Correlation</strong>: Use statistical tools to calculate the correlation coefficient (ranging from -1 to 1) between two stocks. A correlation close to +1 indicates a strong positive relationship.</li><li><strong>Cointegration Test</strong>: Check if the two stocks are cointegrated, meaning they move together in the long run despite short-term deviations.</li><li><strong>Liquidity</strong>: Ensure both stocks are liquid to avoid slippage and ensure smooth execution of trades.</li><li><strong>Volatility</strong>: Look for stocks with similar volatility levels to ensure balanced risk.</li></ol><h3>Potential Pairs for Pair Trading (2014–2023):</h3><h4>1. Banking Sector:</h4><ul><li><strong>HDFC Bank (HDFCBANK) &amp; ICICI Bank (ICICIBANK)</strong></li><li>Both are leading private banks in India and have shown high correlation due to similar exposure to the Indian economy.</li><li><strong>State Bank of India (SBIN) &amp; Bank of Baroda (BANKBARODA)</strong></li><li>Both are public sector banks with similar business models and government influence.</li></ul><h4>2. IT Sector:</h4><ul><li><strong>TCS (TATASTEEL) &amp; Infosys (INFY)</strong></li><li>Both are large-cap IT companies with significant exposure to global markets, leading to high correlation.</li><li><strong>Wipro (WIPRO) &amp; HCL Technologies (HCLTECH)</strong></li><li>Mid-to-large-cap IT firms with similar business models and client bases.</li></ul><h4>3. Automobile Sector:</h4><ul><li><strong>Maruti Suzuki (MARUTI) &amp; Tata Motors (TATAMOTORS)</strong></li><li>Both are major players in the automobile sector, though Tata Motors has a more diversified portfolio.</li><li><strong>Bajaj Auto (BAJAJ-AUTO) &amp; Hero MotoCorp (HEROMOTOCO)</strong></li><li>Both are two-wheeler manufacturers with strong domestic and international presence.</li></ul><h4>4. FMCG Sector:</h4><ul><li><strong>Hindustan Unilever (HINDUNILVR) &amp; ITC (ITC)</strong></li><li>Both are FMCG giants with strong brand presence and distribution networks.</li><li><strong>Nestle India (NESTLEIND) &amp; Britannia Industries (BRITANNIA)</strong></li><li>Both are focused on food products and have shown high correlation.</li></ul><h4>5. Oil &amp; Gas Sector:</h4><ul><li><strong>Reliance Industries (RELIANCE) &amp; ONGC (ONGC)</strong></li><li>Both are major players in the oil and gas sector, though Reliance is more diversified.</li><li><strong>BPCL (BPCL) &amp; HPCL (HINDPETRO)</strong></li><li>Both are public sector oil marketing companies with similar business models.</li></ul><h4>6. Pharma Sector:</h4><ul><li><strong>Sun Pharma (SUNPHARMA) &amp; Dr. Reddy’s Labs (DRREDDY)</strong></li><li>Both are large-cap pharma companies with significant global exposure.</li><li><strong>Cipla (CIPLA) &amp; Lupin (LUPIN)</strong></li><li>Both are mid-to-large-cap pharma companies with similar product portfolios.</li></ul><h3>How to Validate the Pairs:</h3><ol><li><strong>Data Collection</strong>: Gather historical price data for the stocks from 2014 to 2023.</li><li><strong>Correlation Analysis</strong>: Calculate the correlation coefficient using tools like Python, R, or Excel.</li><li><strong>Cointegration Test</strong>: Use statistical tests like the Augmented Dickey-Fuller (ADF) test to check for cointegration.</li><li><strong>Backtesting</strong>: Test the pair trading strategy on historical data to evaluate its performance.</li></ol><h3>Tools for Analysis:</h3><ul><li><strong>Python Libraries</strong>: pandas, numpy, statsmodels, yfinance (for data).</li><li><strong>Platforms</strong>: NSE website, Bloomberg, or other financial data providers.</li></ul><h3>Example of a High-Correlation Pair:</h3><ul><li><strong>HDFC Bank (HDFCBANK) &amp; ICICI Bank (ICICIBANK)</strong>:</li><li>Both stocks have shown a correlation coefficient of over 0.85 in the past decade.</li><li>They are cointegrated, making them suitable for pair trading.</li></ul><h3>Risks to Consider:</h3><ul><li><strong>Sector-Specific Risks</strong>: Economic or regulatory changes can impact the entire sector.</li><li><strong>Divergence Risk</strong>: The correlation may break down due to company-specific events.</li><li><strong>Execution Risk</strong>: Ensure both stocks are liquid to avoid slippage.</li></ul><p>By analyzing the above pairs and validating them with statistical tools, you can identify suitable pairs for pair trading on the NSE. Always backtest your strategy before deploying it in live markets.</p><h4>ChatGPT</h4><p>For pairs trading in the Indian National Stock Exchange (NSE), we can consider stocks from the same sector that are likely to exhibit a high correlation. Here are a few pairs of stocks from different sectors that are known to have strong correlations:</p><h3>IT Sector</h3><ol><li><strong>Tata Consultancy Services (TCS)</strong></li><li><strong>Infosys (INFY)</strong></li></ol><h3>Banking Sector</h3><ol><li><strong>HDFC Bank (HDFCBANK)</strong></li><li><strong>ICICI Bank (ICICIBANK)</strong></li></ol><h3>Automobile Sector</h3><ol><li><strong>Maruti Suzuki (MARUTI)</strong></li><li><strong>Tata Motors (TATAMOTORS)</strong></li></ol><h3>FMCG Sector</h3><ol><li><strong>Hindustan Unilever (HINDUNILVR)</strong></li><li><strong>ITC Limited (ITC)</strong></li></ol><h3>Oil &amp; Gas Sector</h3><ol><li><strong>Reliance Industries (RELIANCE)</strong></li><li><strong>Indian Oil Corporation (IOC)</strong></li></ol><p>For a concrete example, let’s focus on the IT sector with <strong>Tata Consultancy Services (TCS)</strong> and <strong>Infosys (INFY)</strong> as they are large-cap companies with significant market participation and often show correlated movements due to similar business environments.</p><h3>Step-by-Step Implementation</h3><h4>1. Data Collection</h4><p>Collect historical price data for TCS and INFY from 2014 to 2023. You can use sources like Yahoo Finance, NSE’s official website, or financial data APIs like yfinance.</p><h4>2. Cointegration Testing</h4><p>Perform a cointegration test to ensure the chosen pair is suitable for pairs trading.</p><h4>3. Calculating Spread and Z-Score</h4><p>Calculate the spread between the two stocks and standardize it using the Z-Score to determine trading signals.</p><h4>4. Implementing the Trading Strategy</h4><p>Enter trades based on the Z-Score and manage risk with appropriate stop-loss levels.</p><h3>Implementation in Python</h3><p>Here’s a sample implementation using Python:</p><pre>import numpy as np<br>import pandas as pd<br>import yfinance as yf<br>from statsmodels.tsa.stattools import coint<br>import matplotlib.pyplot as plt</pre><pre># Step 1: Fetch historical data<br>start_date = &#39;2014-01-01&#39;<br>end_date = &#39;2023-12-31&#39;<br>tcs = yf.download(&#39;TCS.NS&#39;, start=start_date, end=end_date)<br>infy = yf.download(&#39;INFY.NS&#39;, start=start_date, end=end_date)</pre><pre># Step 2: Prepare the data<br>prices = pd.DataFrame({&#39;TCS&#39;: tcs[&#39;Adj Close&#39;], &#39;INFY&#39;: infy[&#39;Adj Close&#39;]})</pre><pre># Step 3: Cointegration test<br>coint_result = coint(prices[&#39;TCS&#39;], prices[&#39;INFY&#39;])<br>print(f&#39;P-value of cointegration test: {coint_result[1]}&#39;)</pre><pre># Step 4: Calculate the spread and Z-Score<br>prices[&#39;Spread&#39;] = prices[&#39;TCS&#39;] - prices[&#39;INFY&#39;]<br>spread_mean = prices[&#39;Spread&#39;].mean()<br>spread_std = prices[&#39;Spread&#39;].std()<br>prices[&#39;Z-Score&#39;] = (prices[&#39;Spread&#39;] - spread_mean) / spread_std</pre><pre># Step 5: Plotting the Z-Score<br>plt.figure(figsize=(12, 6))<br>plt.plot(prices[&#39;Z-Score&#39;], label=&#39;Z-Score&#39;)<br>plt.axhline(2, color=&#39;r&#39;, linestyle=&#39;--&#39;)<br>plt.axhline(-2, color=&#39;r&#39;, linestyle=&#39;--&#39;)<br>plt.axhline(0, color=&#39;b&#39;, linestyle=&#39;-&#39;)<br>plt.legend()<br>plt.show()</pre><pre># Step 6: Define the trading strategy<br>def signal(data):<br>    buy = []<br>    sell = []<br>    position = 0  # 1 for buy, -1 for sell, 0 for no position</pre><pre>for i in range(len(data)):<br>        if data[&#39;Z-Score&#39;][i] &gt; 2:<br>            if position != -1:<br>                sell.append(data[&#39;TCS&#39;][i])<br>                buy.append(data[&#39;INFY&#39;][i])<br>                position = -1<br>            else:<br>                sell.append(np.nan)<br>                buy.append(np.nan)<br>        elif data[&#39;Z-Score&#39;][i] &lt; -2:<br>            if position != 1:<br>                buy.append(data[&#39;TCS&#39;][i])<br>                sell.append(data[&#39;INFY&#39;][i])<br>                position = 1<br>            else:<br>                sell.append(np.nan)<br>                buy.append(np.nan)<br>        elif abs(data[&#39;Z-Score&#39;][i]) &lt; 0.5:<br>            buy.append(np.nan)<br>            sell.append(np.nan)<br>            position = 0<br>        else:<br>            buy.append(np.nan)<br>            sell.append(np.nan)</pre><pre>return pd.Series(buy, index=data.index), pd.Series(sell, index=data.index)</pre><pre>prices[&#39;Buy&#39;], prices[&#39;Sell&#39;] = signal(prices)</pre><pre># Step 7: Visualize the buy and sell signals<br>plt.figure(figsize=(12, 6))<br>plt.plot(prices[&#39;TCS&#39;], label=&#39;TCS&#39;, alpha=0.5)<br>plt.plot(prices[&#39;INFY&#39;], label=&#39;INFY&#39;, alpha=0.5)<br>plt.scatter(prices.index, prices[&#39;Buy&#39;], label=&#39;Buy Signal&#39;, marker=&#39;^&#39;, color=&#39;g&#39;, alpha=1)<br>plt.scatter(prices.index, prices[&#39;Sell&#39;], label=&#39;Sell Signal&#39;, marker=&#39;v&#39;, color=&#39;r&#39;, alpha=1)<br>plt.legend()<br>plt.show()</pre><p>This script fetches historical price data, performs a cointegration test, calculates the spread and Z-Score, and plots the buy and sell signals. You can adjust the strategy and parameters as needed to suit your specific requirements and market conditions.</p><h4>Gemini</h4><p>I can’t provide specific stock recommendations for pair trading on NSE from 2014–2023 due to regulatory restrictions and the fact that past performance isn’t necessarily indicative of future results. However, I can guide you on how to find such stocks yourself. Here’s what you can do:</p><p><strong>Data Acquisition:</strong></p><ol><li><strong>Focus on F&amp;O Stocks:</strong> Since pair trading often involves shorting one stock, consider focusing on stocks available in the Futures &amp; Options (F&amp;O) segment of the NSE. You can find the F&amp;O list on the NSE website <a href="https://www.nseindia.com/products-services/equity-derivatives-list-underlyings-information">https://www.nseindia.com/products-services/equity-derivatives-list-underlyings-information</a>.</li><li><strong>Download Historical Data:</strong> Several online resources and financial data providers offer historical stock price data for NSE. You’ll need daily closing prices for a ten-year period (2014–2023).</li></ol><p><strong>Identifying High Correlation Pairs:</strong></p><p><strong>Calculate Correlation Coefficient:</strong> Use a tool like Python’s Pandas library or a spreadsheet program like Excel to calculate the correlation coefficient between each pair of stocks in your F&amp;O list. The correlation coefficient measures how closely two stocks move together. A value closer to 1 indicates a strong positive correlation, while a value closer to -1 indicates a strong negative correlation.</p><p><strong>Set Correlation Threshold:</strong> A common threshold for pair trading is 0.8 or higher for a positive correlation. You can adjust this threshold based on your risk tolerance and the number of potential pairs you want to identify.</p><p><strong>Industry and Sector Analysis:</strong> While high correlation is crucial, consider filtering further based on industry or sector. Companies within the same sector often have a natural correlation due to similar market influences.</p><p><strong>Additional Considerations:</strong></p><ul><li><strong>Liquidity:</strong> Ensure both stocks in the pair have sufficient trading volume to enter and exit positions efficiently.</li><li><strong>Trading Costs:</strong> Factor in brokerage fees, shorting costs (if applicable), and other trading expenses when evaluating potential profits.</li><li><strong>Backtesting:</strong> Test your identified pairs using historical data to see how they performed in the past. This helps assess the strategy’s effectiveness but remember, past performance doesn’t guarantee future results.</li></ul><p><strong>Disclaimer:</strong> This information is for educational purposes only and should not be considered financial advice. Always consult with a qualified financial advisor before making any investment decisions.</p><h4>Mistral</h4><p>Pair trading involves identifying two stocks that have a high correlation with each other and then trading them based on the assumption that their relationship will continue. In the Indian NSE, here are a few pairs of stocks that have shown a high correlation over the years and could be considered for pair trading. However, it’s crucial to conduct your own analysis and consider other factors before making any investment decisions.</p><ol><li><strong>Tata Motors and Maruti Suzuki</strong>: Both are major players in the automobile industry. Their correlation could be due to the fact that they are influenced by the same economic and sector-specific factors.</li><li><strong>HDFC Bank and ICICI Bank</strong>: These are two of the largest private sector banks in India. They are likely to be affected by the same regulatory, economic, and industry-specific factors.</li><li><strong>Reliance Industries and BPCL (Bharat Petroleum Corporation Limited)</strong>: Reliance Industries has a significant presence in the petroleum sector, and BPCL is one of the major public sector undertakings in the same sector.</li><li><strong>HUL (Hindustan Unilever Limited) and ITC</strong>: Both are major FMCG (Fast-Moving Consumer Goods) companies in India, and their correlation could be due to their exposure to the same consumer trends and economic factors.</li></ol><h4>Claude:</h4><p>To suggest stocks for pair trading on the Indian National Stock Exchange (NSE) from 2014 to 2023, we need to analyze the correlation between stock pairs during that period. Pair trading involves taking a long position in one stock and a short position in another highly correlated stock to profit from the temporary divergence in their prices.</p><p>Here are a few potential stock pairs that exhibited high correlation during the given period, based on historical data analysis:</p><p>1. Infosys Limited (INFY) and Tata Consultancy Services Limited (TCS)<br> — Both are large-cap IT services companies with a high positive correlation in their stock prices.</p><p>2. Reliance Industries Limited (RELIANCE) and Indian Oil Corporation Limited (IOC)<br> — These are major players in the oil and gas sector, and their stock prices tend to move in tandem due to the nature of the industry.</p><p>3. HDFC Bank Limited (HDFCBANK) and ICICI Bank Limited (ICICIBANK)<br> — As two of the largest private sector banks in India, their stock prices often exhibit a strong positive correlation.</p><p>4. Bharti Airtel Limited (BHARTIARTL) and Vodafone Idea Limited (IDEA)<br> — These are leading telecom service providers, and their stock prices have historically shown a high correlation due to industry dynamics.</p><p>5. Larsen &amp; Toubro Limited (LT) and Bharat Heavy Electricals Limited (BHEL)<br> — These are prominent engineering and construction companies, and their stock prices tend to move together due to similarities in their business segments.</p><p>It’s important to note that correlation can vary over time, and it’s essential to conduct thorough analysis and backtesting before implementing any pair trading strategy. Additionally, factors such as market conditions, company-specific events, and industry trends can impact the correlation between stock pairs.</p><p>While Gemini didn’t suggest any stocks all other models gave us a list of pairs. We will now use equal weight to get the return and risk of each suggestion and do a backtest.</p><h3>Code</h3><pre>import yfinance as yf<br>import pandas as pd<br>import numpy as np<br>import matplotlib.pyplot as plt</pre><pre># Function to calculate performance metrics<br>def calculate_metrics(returns):<br>    cumulative_returns = (1 + returns).cumprod() - 1<br>    sharpe_ratio = returns.mean() / returns.std()<br>    annualized_sharpe_ratio = sharpe_ratio * np.sqrt(252)<br>    drawdown = cumulative_returns - cumulative_returns.cummax()<br>    max_drawdown = drawdown.min()<br>    annualized_returns = (1 + returns.mean()) ** 252 - 1</pre><pre>    return {<br>        &quot;Cumulative Returns&quot;: cumulative_returns.iloc[-1],<br>        &quot;Sharpe Ratio&quot;: sharpe_ratio,<br>        &quot;Annualized Sharpe Ratio&quot;: annualized_sharpe_ratio,<br>        &quot;Maximum Drawdown&quot;: max_drawdown,<br>        &quot;Annualized Returns&quot;: annualized_returns<br>    }</pre><pre># Function to perform pair trading strategy backtest<br>def pair_trading_backtest(pairs, start_date=&quot;2014-01-01&quot;, end_date=&quot;2023-12-31&quot;):<br>    tickers = set([ticker for pair in pairs for ticker in pair])<br>    data = yf.download(list(tickers), start=start_date, end=end_date)[&#39;Adj Close&#39;]<br>    <br>    # Calculate daily returns<br>    returns = data.pct_change().dropna()<br>    <br>    pair_metrics = []<br>    cumulative_returns_list = []</pre><pre>    for stock1, stock2 in pairs:<br>        # Calculate the spread<br>        spread = data[stock1] - data[stock2]<br>        <br>        # Calculate z-score of the spread<br>        zscore = (spread - spread.mean()) / spread.std()<br>        <br>        # Generate trading signals<br>        long_signals = zscore &lt; -1<br>        short_signals = zscore &gt; 1<br>        exit_signals = abs(zscore) &lt; 0.5<br>        <br>        # Initialize positions<br>        positions = pd.DataFrame(index=data.index)<br>        positions[stock1] = 0<br>        positions[stock2] = 0<br>        <br>        # Create positions<br>        positions.loc[long_signals, stock1] = 1<br>        positions.loc[long_signals, stock2] = -1<br>        positions.loc[short_signals, stock1] = -1<br>        positions.loc[short_signals, stock2] = 1<br>        <br>        # Close positions<br>        positions.loc[exit_signals, stock1] = 0<br>        positions.loc[exit_signals, stock2] = 0<br>        <br>        # Forward fill positions<br>        positions = positions.fillna(method=&#39;ffill&#39;)<br>        <br>        # Calculate daily returns of the strategy<br>        daily_returns = (positions.shift(1) * returns).sum(axis=1)<br>        <br>        # Assuming Rs 100 total investment equally distributed<br>        daily_returns = daily_returns / 2<br>        <br>        # Calculate performance metrics<br>        metrics = calculate_metrics(daily_returns)<br>        metrics[&#39;Pair&#39;] = (stock1, stock2)<br>        pair_metrics.append(metrics)<br>        cumulative_returns_list.append((1 + daily_returns).cumprod() - 1)<br>    <br>    # Create DataFrame to display results<br>    results_df = pd.DataFrame(pair_metrics)<br>    <br>    # Calculate average metrics, excluding the &#39;Pair&#39; column<br>    avg_metrics = results_df.drop(columns=[&#39;Pair&#39;]).mean().to_dict()<br>    avg_metrics[&#39;Pair&#39;] = &#39;Average&#39;<br>    avg_metrics_df = pd.DataFrame([avg_metrics])<br>    results_df = pd.concat([results_df, avg_metrics_df], ignore_index=True)</pre><pre>    # Plot cumulative returns for each pair<br>    plt.figure(figsize=(14, 7))<br>    for i, cumulative_returns in enumerate(cumulative_returns_list):<br>        pair = pairs[i]<br>        plt.plot(cumulative_returns, label=f&#39;{pair[0]} &amp; {pair[1]}&#39;)<br>    <br>    plt.title(&#39;Cumulative Returns of Pair Trading Strategy&#39;)<br>    plt.xlabel(&#39;Date&#39;)<br>    plt.ylabel(&#39;Cumulative Returns&#39;)<br>    plt.legend()<br>    plt.show()<br>    <br>    return results_df</pre><pre># Define pairs for each group<br>claude_pairs = [<br>    (&#39;INFY.NS&#39;, &#39;TCS.NS&#39;),<br>    (&#39;RELIANCE.NS&#39;, &#39;IOC.NS&#39;),<br>    (&#39;HDFCBANK.NS&#39;, &#39;ICICIBANK.NS&#39;),<br>    (&#39;BHARTIARTL.NS&#39;, &#39;IDEA.NS&#39;),<br>    (&#39;LT.NS&#39;, &#39;BHEL.NS&#39;)<br>]</pre><pre>mistral_pairs = [<br>    (&#39;TATAMOTORS.NS&#39;, &#39;MARUTI.NS&#39;),<br>    (&#39;HDFCBANK.NS&#39;, &#39;ICICIBANK.NS&#39;),<br>    (&#39;RELIANCE.NS&#39;, &#39;BPCL.NS&#39;),<br>    (&#39;HINDUNILVR.NS&#39;, &#39;ITC.NS&#39;)<br>]</pre><pre>chatgpt_pairs = [<br>    (&#39;TCS.NS&#39;, &#39;INFY.NS&#39;),<br>    (&#39;HDFCBANK.NS&#39;, &#39;ICICIBANK.NS&#39;),<br>    (&#39;MARUTI.NS&#39;, &#39;TATAMOTORS.NS&#39;),<br>    (&#39;HINDUNILVR.NS&#39;, &#39;ITC.NS&#39;),<br>    (&#39;RELIANCE.NS&#39;, &#39;IOC.NS&#39;)<br>]</pre><pre># Run backtest for each group<br>claude_results = pair_trading_backtest(claude_pairs)<br>mistral_results = pair_trading_backtest(mistral_pairs)<br>chatgpt_results = pair_trading_backtest(chatgpt_pairs)</pre><pre># Display results<br>print(&quot;Claude Results:&quot;)<br>print(claude_results)</pre><pre>print(&quot;Mistral Results:&quot;)<br>print(mistral_results)</pre><pre>print(&quot;ChatGPT Results:&quot;)<br>print(chatgpt_results)</pre><h3>Output</h3><h4>Claude Results</h4><pre>Cumulative Returns  Sharpe Ratio  Annualized Sharpe Ratio  \<br>0            0.472290      0.036313                 0.576458   <br>1           -0.222604     -0.010780                -0.171131   <br>2            0.803328      0.048864                 0.775692   <br>3            0.571481      0.025268                 0.401113   <br>4            0.554155      0.032830                 0.521165   <br>5            0.435730      0.026499                 0.420660   </pre><pre>   Maximum Drawdown  Annualized Returns                         Pair  <br>0         -0.189356            0.043129            (INFY.NS, TCS.NS)  <br>1         -0.449846           -0.019165        (RELIANCE.NS, IOC.NS)  <br>2         -0.150183            0.065701  (HDFCBANK.NS, ICICIBANK.NS)  <br>3         -0.638215            0.057203     (BHARTIARTL.NS, IDEA.NS)  <br>4         -0.232151            0.050833             (LT.NS, BHEL.NS)  <br>5         -0.331950            0.039540                      Average</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5s730P_jX3ddJK3ougqoAA.png" /></figure><h4>Mistral Results</h4><pre>Mistral Results:<br>   Cumulative Returns  Sharpe Ratio  Annualized Sharpe Ratio  \<br>0            0.400360      0.024515                 0.389171   <br>1            0.803328      0.048864                 0.775692   <br>2           -0.060937     -0.000725                -0.011504   <br>3            0.758240      0.037862                 0.601045   <br>4            0.475248      0.027629                 0.438601   </pre><pre>   Maximum Drawdown  Annualized Returns                         Pair  <br>0         -0.415828            0.040396   (TATAMOTORS.NS, MARUTI.NS)  <br>1         -0.150183            0.065701  (HDFCBANK.NS, ICICIBANK.NS)  <br>2         -0.348646           -0.001179       (RELIANCE.NS, BPCL.NS)  <br>3         -0.231129            0.065213      (HINDUNILVR.NS, ITC.NS)  <br>4         -0.286446            0.042533                      Average</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QWpPNKW5eV2tLTeTrJAHmQ.png" /></figure><h4>ChatGPT</h4><pre>Cumulative Returns  Sharpe Ratio  Annualized Sharpe Ratio  \<br>0            0.472290      0.036313                 0.576458   <br>1            0.803328      0.048864                 0.775692   <br>2            0.400360      0.024515                 0.389171   <br>3            0.758240      0.037862                 0.601045   <br>4           -0.222604     -0.010780                -0.171131   <br>5            0.442323      0.027355                 0.434247   </pre><pre>   Maximum Drawdown  Annualized Returns                         Pair  <br>0         -0.189356            0.043129            (TCS.NS, INFY.NS)  <br>1         -0.150183            0.065701  (HDFCBANK.NS, ICICIBANK.NS)  <br>2         -0.415828            0.040396   (MARUTI.NS, TATAMOTORS.NS)  <br>3         -0.231129            0.065213      (HINDUNILVR.NS, ITC.NS)  <br>4         -0.449846           -0.019165        (RELIANCE.NS, IOC.NS)  <br>5         -0.287268            0.039055                      Average</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gA2clFYcZIUlTiz3_BDs0Q.png" /></figure><h4>DeepSeek Result</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7Fe7fA-Nb22Fjd_wgzYECw.png" /></figure><pre>    Maximum Drawdown  Annualized Returns                            Pair  <br>0          -0.150183            0.069030     (HDFCBANK.NS, ICICIBANK.NS)  <br>1          -0.261374            0.041348        (SBIN.NS, BANKBARODA.NS)  <br>2          -0.146087            0.043486               (TCS.NS, INFY.NS)  <br>3          -0.150732           -0.002013          (WIPRO.NS, HCLTECH.NS)  <br>4          -0.345952            0.037871      (MARUTI.NS, TATAMOTORS.NS)  <br>5          -0.160318           -0.002728  (BAJAJ-AUTO.NS, HEROMOTOCO.NS)  <br>6          -0.229066            0.066526         (HINDUNILVR.NS, ITC.NS)  <br>7          -0.280824            0.065910    (NESTLEIND.NS, BRITANNIA.NS)  <br>8          -0.264546            0.060213          (RELIANCE.NS, ONGC.NS)  <br>9          -0.190672            0.028519         (BPCL.NS, HINDPETRO.NS)  <br>10         -0.266601            0.030458      (SUNPHARMA.NS, DRREDDY.NS)  <br>11         -0.277367            0.006255            (CIPLA.NS, LUPIN.NS)  <br>12         -0.226977            0.037073                         Average</pre><pre>    Cumulative Returns  Sharpe Ratio  Annualized Sharpe Ratio  <br>0             0.979735      0.052116                 0.827320   <br>1             0.501146      0.034524                 0.548047   <br>2             0.535148      0.036367                 0.577301   <br>3            -0.030923     -0.002985                -0.047387   <br>4             0.417651      0.024109                 0.382718   <br>5            -0.041418     -0.003510                -0.055712   <br>6             0.895923      0.040751                 0.646908   <br>7             0.921853      0.051108                 0.811315   <br>8             0.747102      0.032051                 0.508788   <br>9             0.319047      0.025693                 0.407859   <br>10            0.329892      0.022585                 0.358522   <br>11            0.022158      0.004285                 0.068019   <br>12            0.466443      0.026424                 0.419475   </pre><p><strong>Average Cumulative Returns</strong>:</p><ul><li><strong>Mistral</strong> has the highest average cumulative returns (0.475248), indicating that the pairs in this group performed the best overall.</li><li><strong>DeepSeek</strong> has the second highest cumulative returns ( 0.466443).</li><li><strong>ChatGPT</strong> follows with an average cumulative return of (0.442323).</li><li><strong>Claude</strong> has the lowest average cumulative return (0.435730).</li></ul><p><strong>Average Sharpe Ratio</strong>:</p><ul><li><strong>Mistral</strong> also leads with the highest average Sharpe ratio (0.027629), which suggests it has the best risk-adjusted return among the three groups.</li><li><strong>ChatGPT</strong> is slightly behind with an average Sharpe ratio of 0.027355.</li><li><strong>Claude</strong> has the second lowest average Sharpe ratio (0.026499).</li><li><strong>Deep Seek</strong> pairs come next with an average Sharpe ratio of 0.026424, showing decent risk-adjusted returns.</li></ul><p><strong>Average Annualized Sharpe Ratio</strong>:</p><ul><li><strong>Mistral</strong> has the highest average annualized Sharpe ratio (0.438601), indicating strong risk-adjusted performance on an annual basis.</li><li><strong>ChatGPT</strong> follows with an average annualized Sharpe ratio of 0.434247.</li><li><strong>Claude</strong> has the Second lowest annualized Sharpe ratio (0.420660).</li><li>Deep Seek pairs have an average annualized Sharpe ratio of <strong>0.419475</strong>,</li></ul><p><strong>Average Maximum Drawdown</strong>:</p><ul><li><strong>Claude</strong> has the highest maximum drawdown (-0.331950), indicating higher risk and larger peak-to-trough declines in portfolio value.</li><li><strong>ChatGPT</strong> has a lower average maximum drawdown (-0.287268), suggesting better risk control.</li><li><strong>Mistral</strong> has the Second lowest maximum drawdown (-0.286446), indicating it experienced the smallest peak-to-trough declines and managed risk better compared to the other groups.</li><li>Deep Seek pairs experienced lowest drawdown with an average maximum drawdown of -0.226977,</li></ul><p><strong>Average Annualized Returns</strong>:</p><ul><li><strong>Mistral</strong> leads with the highest average annualized returns (0.042533), indicating it provided the best returns on an annual basis.</li><li><strong>Claude</strong> follows with average annualized returns of 0.039540.</li><li><strong>ChatGPT</strong> has the lowest average annualized returns (0.039055).</li><li><strong>Deep Seek </strong>pairs achieved lowest an average annualized return of 0.037073</li></ul><h3>Overall Summary:</h3><ul><li><strong>Mistral</strong> performed the best overall, with the highest cumulative returns, Sharpe ratio, annualized Sharpe ratio, and annualized returns, while also managing risk effectively with the lowest maximum drawdown.</li><li><strong>Deep Seek </strong>pairs showed strong overall performance, ranking second in cumulative returns and providing consistent risk-adjusted metrics with moderate drawdowns.</li><li><strong>ChatGPT</strong> also showed strong performance, especially in risk-adjusted returns, and maintained relatively low drawdown.</li><li><strong>Claude</strong>, while still profitable, lagged behind Mistral and ChatGPT in most performance metrics and had higher risk as indicated by the maximum drawdown.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3de9b3d068d0" width="1" height="1" alt=""><hr><p><a href="https://levelup.gitconnected.com/is-deepseek-better-than-chatgpt-claude-mistral-and-gemini-for-pair-trading-3de9b3d068d0">Is DeepSeek Better Than ChatGPT , Claude, Mistral, and Gemini for Pair Trading ?</a> was originally published in <a href="https://levelup.gitconnected.com/">Level Up Coding</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How AI Agents Can Transform Manufacturing Plants]]></title>
            <link>https://levelup.gitconnected.com/how-ai-agents-can-transform-manufacturing-plants-0cb1a0942d54?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/0cb1a0942d54</guid>
            <category><![CDATA[python]]></category>
            <category><![CDATA[manufacturing]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[ai-agent]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Mon, 27 Jan 2025 02:59:19 GMT</pubDate>
            <atom:updated>2025-01-27T02:59:19.880Z</atom:updated>
            <content:encoded><![CDATA[<h4>A Step-by-Step Guide</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NHgsCerQCmS34SxeRhfSIQ.png" /><figcaption>Dalle Generated</figcaption></figure><p>A <strong>Digital Twin</strong> is a virtual representation of a physical object, system, or process that simulates real-time operations and behaviors. It integrates data from sensors and machines, creating a dynamic, real-time model that mirrors the performance of its physical counterpart. This concept has rapidly gained momentum, especially in industries like manufacturing, where it is used for predictive maintenance, performance optimization, and process monitoring.</p><p>In the context of manufacturing, Digital Twins can represent individual machines, entire production lines, or even the whole plant. By capturing sensor data — such as temperature, pressure, humidity, and speed — Digital Twins provide a live, digital reflection of a physical system. This enables remote monitoring and allows operators to make real-time adjustments, improving efficiency and identifying issues before they escalate into significant problems. For instance, if a temperature sensor detects an abnormal rise in temperature, the Digital Twin can alert operators, enabling timely intervention and preventing potential machinery damage or downtime.</p><h3>The Role of AI Agents in Digital Twin Monitoring</h3><p>While Digital Twins provide a digital mirror of physical systems, <strong>AI Agents</strong> enhance their capabilities by enabling autonomous monitoring, real-time analysis, and intelligent anomaly detection. AI agents are software entities designed to carry out specific tasks autonomously by processing data, making decisions, and interacting with external tools or systems.</p><p>For example, in a manufacturing plant’s Digital Twin environment, AI agents can monitor sensor data from machinery. If the data shows an anomaly — like an unexpectedly high temperature — the agent can flag this as a potential problem and trigger corrective actions, such as logging the issue in a database or alerting operators.</p><p>By leveraging AI agents, manufacturers can go beyond traditional monitoring systems. AI agents allow for smarter, automated decision-making that improves system reliability and reduces human error.</p><h3>Building AI Agents to Monitor and Log Discrepancies</h3><p>To better understand how AI agents work within the Digital Twin environment, let’s explore how we can build an AI-based solution to monitor sensor readings and log any discrepancies. In this scenario, we are using <strong>PydanticAI</strong>, a Python agent framework that ensures type safety and simplifies the integration of generative models like OpenAI’s GPT. The AI agents will monitor the sensor data, detect discrepancies, and log them into a PostgreSQL database for further analysis.</p><p>We will implement two types of AI agents:</p><ol><li><strong>Monitoring Agent</strong>: This agent will check sensor data for discrepancies, such as deviations from expected temperature or pressure levels.</li><li><strong>Logging Agent</strong>: When the Monitoring Agent identifies a discrepancy, the Logging Agent will record the details in a database for future reference.</li></ol><h3>Example Scenario: Manufacturing Plant Monitoring</h3><p>Imagine a manufacturing plant where machines are equipped with sensors to monitor variables such as temperature, pressure, and speed. These sensors send data to a <strong>Digital Twin</strong>, which reflects the real-time state of the machinery.</p><p>The <strong>Monitoring Agent</strong> will receive sensor data, such as:</p><ul><li><strong>Sensor Type</strong>: Temperature</li><li><strong>Expected Value</strong>: 70°F</li><li><strong>Actual Value</strong>: 75°F</li></ul><p>The agent will compare the expected value with the actual reading. If a significant discrepancy is detected, such as the temperature exceeding the expected threshold, the agent will flag this as a potential issue, e.g., <strong>“High Temperature”</strong>. The <strong>Logging Agent</strong> will then store this information in the PostgreSQL database, making it available for further investigation and action.</p><h4>How the AI Agents Work in the Digital Twin System</h4><ol><li><strong>Sensor Data Input</strong>: The AI agents receive sensor data from the Digital Twin in real-time.</li><li><strong>Discrepancy Detection</strong>: The <strong>Monitoring Agent</strong> compares actual sensor readings with expected values. If discrepancies exist (e.g., the temperature exceeds a predefined threshold), the agent flags this.</li><li><strong>Log Discrepancy</strong>: If a discrepancy is detected, the <strong>Logging Agent</strong> logs it into a PostgreSQL database, ensuring a record is available for future analysis or corrective actions.</li></ol><h3>The Business Problem it Solves</h3><p>In a manufacturing environment, machine failures, inefficiencies, and downtime are costly. A single, undetected issue — such as an overheating machine — can lead to system breakdowns, unplanned downtime, and significant repair costs. Timely detection of such issues is critical to maintaining productivity and minimizing disruptions.</p><p>This is where the AI agent-based solution comes in. By monitoring sensor data and detecting discrepancies in real-time, AI agents can solve several key business problems:</p><ol><li><strong>Preventing Downtime</strong>: By detecting anomalies early, AI agents can prevent costly machine breakdowns, thus reducing unplanned downtime. Early intervention can save companies from expensive repairs and production delays.</li><li><strong>Improved Data Handling and Analysis</strong>: Traditional monitoring systems often rely on manual checks and logs. AI agents can automate the analysis of large volumes of sensor data in real-time, offering quicker and more accurate detection of discrepancies.</li><li><strong>Automated Monitoring and Logging</strong>: Automating the monitoring and logging of sensor discrepancies reduces the need for manual intervention, which allows plant operators to focus on more critical tasks. Additionally, automation ensures that no discrepancies are overlooked.</li><li><strong>Better Decision Making</strong>: With a comprehensive record of all discrepancies stored in the database, plant managers can analyze patterns, optimize workflows, and make data-driven decisions about maintenance schedules, resource allocation, and process improvements.</li><li><strong>Predictive Maintenance</strong>: By logging discrepancies over time, manufacturers can adopt a predictive maintenance approach. For example, if temperature anomalies are detected consistently over time, the plant can schedule preventive maintenance, reducing the likelihood of unexpected failures and prolonging equipment life.</li></ol><h3>Code Walkthrough</h3><h4>1. Set Up PostgreSQL Database</h4><p>We begin by creating a table in PostgreSQL to store discrepancies detected by the AI agents. This table will include fields for the sensor type, expected value, actual value, discrepancy type, and the timestamp of when the discrepancy occurred.</p><pre>CREATE TABLE IF NOT EXISTS sensor_discrepancies (<br>    id SERIAL PRIMARY KEY,<br>    timestamp TIMESTAMP DEFAULT CURRENT_TIMESTAMP,<br>    sensor_type VARCHAR(50),<br>    expected_value FLOAT,<br>    actual_value FLOAT,<br>    discrepancy_type VARCHAR(50)<br>);</pre><h4>2. Install Dependencies</h4><p>Next, install the required dependencies for the project:</p><pre>pip install pydantic-ai psycopg2 asyncpg openai</pre><h4>3. Creating the AI Agents</h4><p>The system consists of two agents: the <strong>Monitoring Agent</strong> and the <strong>Logging Agent</strong>.</p><h4>1. Monitoring Agent</h4><p>The <strong>Monitoring Agent</strong> analyzes the incoming sensor data and checks for discrepancies. If the actual sensor readings are outside the expected range, the agent flags them as discrepancies.</p><pre>from pydantic_ai import Agent<br>from pydantic_ai.models.openai import OpenAIModel<br>from pydantic import BaseModel</pre><pre># Define the SensorDiscrepancy model<br>class SensorDiscrepancy(BaseModel):<br>    sensor_type: str<br>    expected_value: float<br>    actual_value: float<br>    discrepancy_type: str</pre><pre># Create the monitoring agent<br>monitoring_model = OpenAIModel(&#39;gpt-4&#39;, api_key=&quot;your-openai-api-key&quot;)<br>monitoring_agent = Agent(<br>    monitoring_model,<br>    result_type=SensorDiscrepancy,<br>    system_prompt=(<br>        &quot;You are a sensor monitoring assistant for a manufacturing plant. &quot;<br>        &quot;Monitor sensor readings for discrepancies. If a sensor reading is outside the expected range, &quot;<br>        &quot;log the discrepancy in a structured format with sensor type, expected value, actual value, and the type of discrepancy.&quot;<br>    )<br>)</pre><h4>2. Logging Agent</h4><p>The <strong>Logging Agent</strong> logs any discrepancies found by the Monitoring Agent into the PostgreSQL database for future analysis.</p><pre>import asyncpg</pre><pre>class DatabaseConn:<br>    def __init__(self):<br>        self.dsn = &quot;your-postgresql-connection-string-here&quot;</pre><pre>    async def _connect(self):<br>        return await asyncpg.connect(self.dsn)</pre><pre>    async def log_discrepancy(self, sensor_type, expected_value, actual_value, discrepancy_type):<br>        query = &quot;&quot;&quot;<br>        INSERT INTO sensor_discrepancies (sensor_type, expected_value, actual_value, discrepancy_type)<br>        VALUES ($1, $2, $3, $4);<br>        &quot;&quot;&quot;<br>        conn = await self._connect()<br>        try:<br>            result = await conn.execute(query, sensor_type, expected_value, actual_value, discrepancy_type)<br>            return result == &quot;INSERT 0 1&quot;<br>        finally:<br>            await conn.close()</pre><h4>Discrepancy Detection and Logging</h4><p>The agents are then wired together to detect discrepancies and log them in the database.</p><pre>async def monitor_and_log_sensor_data(sensor_data: dict):<br>    # Run the monitoring agent to check for discrepancies<br>    monitoring_result = await monitoring_agent.run(sensor_data)</pre><pre>    if monitoring_result.data.discrepancy_type != &quot;None&quot;:<br>        # If discrepancy is found, log it using the logging agent<br>        logging_result = await logging_agent.run(<br>            sensor_type=monitoring_result.data.sensor_type,<br>            expected_value=monitoring_result.data.expected_value,<br>            actual_value=monitoring_result.data.actual_value,<br>            discrepancy_type=monitoring_result.data.discrepancy_type<br>        )<br>        print(logging_result)</pre><h4>Simulated Sensor Data and Execution</h4><p>Finally, simulate some sensor data to test the system:</p><pre>sensor_data = {<br>    &quot;sensor_type&quot;: &quot;Temperature&quot;,<br>    &quot;expected_value&quot;: 70.0,<br>    &quot;actual_value&quot;: 75.0,  # Discrepancy here<br>    &quot;discrepancy_type&quot;: &quot;High&quot;<br>}</pre><pre>import asyncio<br>asyncio.run(monitor_and_log_sensor_data(sensor_data))</pre><h3>4. Conclusion</h3><p>By integrating <strong>AI Agents</strong> into a <strong>Digital Twin</strong> environment, manufacturers can achieve smarter, more efficient monitoring of sensor data. The <strong>Monitoring Agent</strong> detects discrepancies in real-time, while the <strong>Logging Agent</strong> records these discrepancies in a database for further analysis. This automated system helps prevent downtime, improves decision-making, and supports predictive maintenance strategies, ultimately leading to a more optimized and cost-efficient manufacturing process.</p><p>With AI and Digital Twin technology working together, manufacturers can stay ahead of potential issues, reduce costly disruptions, and enhance the overall reliability of their operations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0cb1a0942d54" width="1" height="1" alt=""><hr><p><a href="https://levelup.gitconnected.com/how-ai-agents-can-transform-manufacturing-plants-0cb1a0942d54">How AI Agents Can Transform Manufacturing Plants</a> was originally published in <a href="https://levelup.gitconnected.com/">Level Up Coding</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What is Fine-Tuning? How to Make AI Work for You]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://levelup.gitconnected.com/what-is-fine-tuning-how-to-make-ai-work-for-you-8304772f9657?source=rss-8ed1df361839------2"><img src="https://cdn-images-1.medium.com/max/1024/1*NI3btAoHiTiB1D4cRtZfSA.png" width="1024"></a></p><p class="medium-feed-snippet">It is fine-tuning when you take something that already is working and then make it even better for your needs. This is akin to buying a&#x2026;</p><p class="medium-feed-link"><a href="https://levelup.gitconnected.com/what-is-fine-tuning-how-to-make-ai-work-for-you-8304772f9657?source=rss-8ed1df361839------2">Continue reading on Level Up Coding »</a></p></div>]]></description>
            <link>https://levelup.gitconnected.com/what-is-fine-tuning-how-to-make-ai-work-for-you-8304772f9657?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/8304772f9657</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[gpt]]></category>
            <category><![CDATA[fine-tuning]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Mon, 21 Oct 2024 21:24:13 GMT</pubDate>
            <atom:updated>2024-10-21T21:24:13.386Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[How to Bulk Add/Edit Alt Text to Images in Shopify Using AltText AI Bulk Generator SEO+]]></title>
            <link>https://medium.com/sageskills-technology-shopify-app-developer/how-to-add-alt-text-to-images-in-shopify-using-alttext-ai-bulk-generator-seo-cac40f16d04d?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/cac40f16d04d</guid>
            <category><![CDATA[image]]></category>
            <category><![CDATA[seo]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[shopify]]></category>
            <category><![CDATA[alt-text]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Mon, 14 Oct 2024 10:01:59 GMT</pubDate>
            <atom:updated>2024-10-14T10:03:18.568Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xo07YdKMXjLcvaIb" /></figure><p>As a Shopify merchant, you understand the importance of product visibility and accessibility. One simple yet often overlooked way to improve both is by adding alt text to your images. Alt text, or “alternative text,” is a brief description of an image that helps search engines understand what the image represents. Not only does it enhance your SEO, but it also improves your store’s accessibility, making it easier for visually impaired customers to navigate.</p><p>However, writing alt text for each image can be a time-consuming and tedious task. With hundreds or even thousands of product images, the workload adds up quickly. This is where <em>AltText AI Bulk Generator SEO+</em> can revolutionize your workflow.</p><h3>Why Alt Text Matters for SEO and Accessibility</h3><p>Alt text serves two critical purposes:</p><ol><li><strong>SEO Benefits</strong>: Search engines rely on alt text to index images, making them more discoverable. Well-written, SEO-friendly alt text can boost your page rankings and drive organic traffic to your store.</li><li><strong>Accessibility</strong>: For customers using screen readers, alt text helps describe what is shown in an image, enhancing their shopping experience.</li></ol><p>Without optimized alt text, you may miss out on valuable opportunities for organic growth and risk alienating a portion of your customer base.</p><h3>Does Alt Text Show Up on Shopify?</h3><p>Yes, alt text is visible on Shopify, but only in specific contexts. While alt text itself is not shown directly on your store’s pages, it is embedded in the HTML of your product images. This means that customers won’t see the alt text when browsing your store, but search engines like Google will read it to index your images properly. Additionally, screen readers will use the alt text to describe images to visually impaired customers, improving your store’s accessibility.</p><p>To view or edit alt text manually in Shopify, you can go to your product or media settings, click on the image, and add or update the alt text field. However, this process can be time-consuming, especially if you have a large inventory. That’s where <em>AltText AI Bulk Generator SEO+</em> comes in, making it easier to manage alt text in bulk and ensure your images are optimized for both SEO and accessibility with minimal effort.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*rcmx2w3NlWKdj6ic" /></figure><h3>How AltText AI Bulk Generator SEO+ Solves This Problem</h3><p>Link: <a href="https://apps.shopify.com/alt-text-seo-gen-ai">https://apps.shopify.com/alt-text-seo-gen-ai</a></p><p>Writing consistent, SEO-friendly alt text can be a struggle. That’s why we created <em>AltText AI Bulk Generator SEO+</em>, a tool designed to simplify and automate the alt text creation process. With a single click, you can generate alt text for all your images, saving you hours of work and ensuring your store is optimized for search engines and accessible to everyone.</p><p><strong>Key Features of AltText AI Bulk Generator SEO+:</strong></p><ul><li><strong>Accessibility Support</strong>: Automatically generates descriptive alt text for every image, helping you meet accessibility guidelines and ensure a better shopping experience for all customers.</li><li><strong>One-Click Bulk Generation</strong>: Instead of writing alt text for each image individually, generate alt text for your entire store in seconds.</li><li><strong>SEO Optimization</strong>: Our AI creates SEO-friendly alt text, helping you improve your store’s ranking on search engines.</li><li><strong>Customizable Output</strong>: You can edit and refine the AI-generated alt text to match your store’s tone of voice or branding.</li><li><strong>Effortless Integration</strong>: Seamlessly connect the app to your Shopify store in just minutes, making it easy to start improving your SEO immediately.</li></ul><h3>Step-by-Step Guide to Adding Alt Text in Shopify Using AltText AI Bulk Generator SEO+</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FRxMjHXWbqB8%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DRxMjHXWbqB8&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FRxMjHXWbqB8%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/bc2933f3793ced23de4b1f80b10bb0db/href">https://medium.com/media/bc2933f3793ced23de4b1f80b10bb0db/href</a></iframe><ol><li><strong>Install the App</strong><br>Begin by downloading and installing <em>AltText AI Bulk Generator SEO+</em> from the Shopify App Store <a href="https://apps.shopify.com/alt-text-seo-gen-ai">here</a>. The installation is simple and only takes a few minutes.</li><li><strong>Connect Your Store</strong><br>Once installed, connect the app to your Shopify store. The app will automatically pull your product images into its dashboard.</li><li><strong>Bulk Generate Alt Text</strong><br>With just one click, the app will scan all your images and generate descriptive, SEO-optimized alt text for each one. You’ll have the option to review and edit the text to ensure it aligns with your brand voice.</li><li><strong>Apply Changes</strong><br>After generating and reviewing the alt text, apply the changes to your Shopify store. The updated alt text will now be visible to search engines and customers using screen readers.</li><li><strong>Monitor and Improve SEO</strong><br>Over time, you’ll see improvements in your store’s SEO performance. With optimized alt text, your products will rank higher in search results, leading to more organic traffic and potential sales.</li></ol><h3>Conclusion</h3><p>Adding alt text to your product images is essential for improving both SEO and accessibility, but it doesn’t have to be a chore. With <em>AltText AI Bulk Generator SEO+</em>, you can streamline the process and ensure your store is always optimized, accessible, and ready to rank higher in search engine results. Try it today and experience the benefits of automated alt text generation.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cac40f16d04d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/sageskills-technology-shopify-app-developer/how-to-add-alt-text-to-images-in-shopify-using-alttext-ai-bulk-generator-seo-cac40f16d04d">How to Bulk Add/Edit Alt Text to Images in Shopify Using AltText AI Bulk Generator SEO+</a> was originally published in <a href="https://medium.com/sageskills-technology-shopify-app-developer">Sageskills Technology Shopify App Developer</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[4 Advanced Prompt techniques to take your productivity to next level]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://levelup.gitconnected.com/4-advanced-prompt-techniques-to-take-your-productivity-to-next-level-35042bbdfa2b?source=rss-8ed1df361839------2"><img src="https://cdn-images-1.medium.com/max/1024/1*cx3u1Bt3rKjJQGL4-Ltxbw.jpeg" width="1024"></a></p><p class="medium-feed-snippet">Unleash the superpower within</p><p class="medium-feed-link"><a href="https://levelup.gitconnected.com/4-advanced-prompt-techniques-to-take-your-productivity-to-next-level-35042bbdfa2b?source=rss-8ed1df361839------2">Continue reading on Level Up Coding »</a></p></div>]]></description>
            <link>https://levelup.gitconnected.com/4-advanced-prompt-techniques-to-take-your-productivity-to-next-level-35042bbdfa2b?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/35042bbdfa2b</guid>
            <category><![CDATA[inspiration]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[prompt-engineering]]></category>
            <category><![CDATA[creativity]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Wed, 04 Sep 2024 09:51:48 GMT</pubDate>
            <atom:updated>2024-09-04T09:51:48.735Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[What is Prompt Chaining, and How to Use It?]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://levelup.gitconnected.com/what-is-prompt-chaining-and-how-to-use-it-2f4311cf0b44?source=rss-8ed1df361839------2"><img src="https://cdn-images-1.medium.com/max/1024/1*zY08Ze-22etioBODIhRXkg.jpeg" width="1024"></a></p><p class="medium-feed-snippet">Ever tried assembling complicated furniture without instructions? Well, if yes, you probably know how something important can get missed&#x2026;</p><p class="medium-feed-link"><a href="https://levelup.gitconnected.com/what-is-prompt-chaining-and-how-to-use-it-2f4311cf0b44?source=rss-8ed1df361839------2">Continue reading on Level Up Coding »</a></p></div>]]></description>
            <link>https://levelup.gitconnected.com/what-is-prompt-chaining-and-how-to-use-it-2f4311cf0b44?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/2f4311cf0b44</guid>
            <category><![CDATA[prompt]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Mon, 02 Sep 2024 01:02:30 GMT</pubDate>
            <atom:updated>2024-09-02T01:02:30.270Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[What are AI Agents, and how are they built?]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://levelup.gitconnected.com/what-are-ai-agents-and-how-are-they-built-4ae0fe5d5203?source=rss-8ed1df361839------2"><img src="https://cdn-images-1.medium.com/max/1024/1*fUMTcqii4A0gak1wruReEw.jpeg" width="1024"></a></p><p class="medium-feed-snippet">React is the secret ingredient!</p><p class="medium-feed-link"><a href="https://levelup.gitconnected.com/what-are-ai-agents-and-how-are-they-built-4ae0fe5d5203?source=rss-8ed1df361839------2">Continue reading on Level Up Coding »</a></p></div>]]></description>
            <link>https://levelup.gitconnected.com/what-are-ai-agents-and-how-are-they-built-4ae0fe5d5203?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/4ae0fe5d5203</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[react]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[automation]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Mon, 02 Sep 2024 01:02:22 GMT</pubDate>
            <atom:updated>2024-09-02T01:02:22.536Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[3 Powerful Use Cases of AI Beyond Chatbots]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://levelup.gitconnected.com/3-powerful-use-cases-of-ai-beyond-chatbots-ab4d0653d014?source=rss-8ed1df361839------2"><img src="https://cdn-images-1.medium.com/max/1024/1*sSHugE9l2dgm7G5RhdUK8Q.png" width="1024"></a></p><p class="medium-feed-snippet">As Generative AI gains popularity by the day, it is only natural for chatbots to hog the limelight. Be it chatbots for customer service or&#x2026;</p><p class="medium-feed-link"><a href="https://levelup.gitconnected.com/3-powerful-use-cases-of-ai-beyond-chatbots-ab4d0653d014?source=rss-8ed1df361839------2">Continue reading on Level Up Coding »</a></p></div>]]></description>
            <link>https://levelup.gitconnected.com/3-powerful-use-cases-of-ai-beyond-chatbots-ab4d0653d014?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/ab4d0653d014</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[python]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Tue, 27 Aug 2024 16:35:38 GMT</pubDate>
            <atom:updated>2024-08-27T16:35:38.765Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[5 Essential Python Libraries for Data Analysis in Finance]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://levelup.gitconnected.com/5-essential-python-libraries-for-data-analysis-in-finance-c3494032e32d?source=rss-8ed1df361839------2"><img src="https://cdn-images-1.medium.com/max/1024/1*6_ZjRF571jDT5UICMAuHKg.png" width="1024"></a></p><p class="medium-feed-snippet">In the very dynamic finance world, it becomes central to make sound decisions based on analysis. Python opens the way for a rich ecosystem&#x2026;</p><p class="medium-feed-link"><a href="https://levelup.gitconnected.com/5-essential-python-libraries-for-data-analysis-in-finance-c3494032e32d?source=rss-8ed1df361839------2">Continue reading on Level Up Coding »</a></p></div>]]></description>
            <link>https://levelup.gitconnected.com/5-essential-python-libraries-for-data-analysis-in-finance-c3494032e32d?source=rss-8ed1df361839------2</link>
            <guid isPermaLink="false">https://medium.com/p/c3494032e32d</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[numpy]]></category>
            <category><![CDATA[finance]]></category>
            <dc:creator><![CDATA[RAHULA RAJ]]></dc:creator>
            <pubDate>Fri, 23 Aug 2024 11:33:55 GMT</pubDate>
            <atom:updated>2024-08-23T11:33:55.964Z</atom:updated>
        </item>
    </channel>
</rss>