<?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 Pawan Wagh on Medium]]></title>
        <description><![CDATA[Stories by Pawan Wagh on Medium]]></description>
        <link>https://medium.com/@pwnwagh?source=rss-21cbaa373d88------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*rI8Xqn1oQiWXerWPUyXi2w.jpeg</url>
            <title>Stories by Pawan Wagh on Medium</title>
            <link>https://medium.com/@pwnwagh?source=rss-21cbaa373d88------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 20:17:33 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@pwnwagh/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[Building & Deploying Strategies: Wizzer’s Code Editor & Container Infrastructure]]></title>
            <link>https://medium.com/wizzercounsel/building-deploying-strategies-wizzers-code-editor-container-infrastructure-7386a1705fcf?source=rss-21cbaa373d88------2</link>
            <guid isPermaLink="false">https://medium.com/p/7386a1705fcf</guid>
            <category><![CDATA[wizzer-tech]]></category>
            <dc:creator><![CDATA[Pawan Wagh]]></dc:creator>
            <pubDate>Mon, 31 Mar 2025 05:12:06 GMT</pubDate>
            <atom:updated>2025-03-31T05:12:06.134Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fuWlPuQUsNaAEKybhE2AAg.png" /></figure><p>Our first post focused on <strong>data pipelines</strong>, this second post of our Wizzer Engineering Vision series deep dives into <strong>infrastructure and the code editor</strong> that let traders <strong>write, backtest, and deploy</strong> their strategies. We’ve built a developer-centric workflow — complete with version control, container isolation, and integrated CI/CD pipelines — that packages &amp; deploys the strategies as secure &amp; isolated containers.</p><h3>Why We Built a Code Editor for Traders?</h3><p>Most traders and quants have become used to either custom scripts or rigid “black-box” systems. We wanted bring something similar with modern software development — familiar tooling, version control, and a smooth deployment process.</p><p>So we took the same underlying editor engine that powers VSCode and embedded it into our trading terminal. This means:</p><ul><li><strong>Syntax Highlighting &amp; IntelliSense</strong>: You get quick hints and error checks.</li><li><strong>Version Control Built-In</strong>: You can commit your code changes without leaving the browser.</li><li><strong>Ease of Use</strong>: If you’re used to VSCode shortcuts and layouts, you’ll feel right at home.</li></ul><p>Essentially, we treat your trading strategy like a real piece of software — because it is.</p><h3>Each Strategy as a Git Repository</h3><p>When you first create a new strategy, we create a Git repository just for that strategy. Why Git? Because <strong>version control</strong> is a game-changer for algo trading:</p><ol><li><strong>Version Tracking</strong>: If you break something with a new commit, you can always roll back to the old version. Now, you can get all of these questions answered — Who changed what? When it got changed? Can we the see the historical log of the exactly what was changed before this version was deployed? and, a lot more.</li><li><strong>Team Collaboration</strong>: If you work with others, they can review your code or propose changes.</li><li><strong>CI/CD Integration</strong>: Once your code’s in Git, we can automate backtesting, building, and deployment.</li></ol><p>No more “I remember a script from last month that did X.”, or “The market has sentiment has changed, let’s deploy ‘that’ of the strategy”. Now you know exactly where everything lives — and how it’s changed over time. Everything is versioned and trackable.</p><h3>One-Click Deployment: The CI/CD Pipeline</h3><p>After you commit your code, you’ll see a <strong>Deploy</strong> button in our trading terminal. One click, and our CI/CD pipeline gets into action. Here’s the high-level flow:</p><ol><li><strong>Trigger</strong>: Your commit notifies our pipeline.</li><li><strong>Build</strong>: We bundle the strategy into a Docker image. This image includes the Wizzer SDK(s) and only the libraries we’ve whitelisted to keep things safe.</li><li><strong>Sandboxing</strong>: We spin up an ephemeral container to build and validate your strategy code.</li><li><strong>Deployment</strong>: If everything checks out, we push the Docker image to our container infrastructure, along with any user/strategy metadata so it can connect to live quotes and data.</li><li><strong>Logs &amp; Feedback</strong>: You can watch all the logs in real time — packaging steps, build logs, final deployment status. If something fails, you’ll know exactly why.</li></ol><p>It’s basically the <strong>“push to production”</strong> flow that software developers enjoy, but adapted for algo trading.</p><h3>Container Infrastructure (The Sandbox Approach)</h3><p>Each strategy runs in its own Docker container, fully isolated from other strategies. That isolation is crucial for:</p><ul><li><strong>Security</strong>: Your code can’t tamper with someone else’s code.</li><li><strong>Reliability</strong>: If your container goes down, it doesn’t affect the rest of the platform.</li><li><strong>Easy Updates</strong>: When you commit new code, we replace the old container with a fresh one. No messy config issues.</li><li><strong>Resource optimization: </strong>Containerization allows us to allocate &amp; limit resources — CPU, memory, etc — , so no strategy can affect the performance of other strategies.</li><li><strong>Reproducibility: </strong>Strategy as containers solves another problem — multiple deployments of the same or different versions of the strategy will always ensure the environment — OS, libraries, dependencies— is reproducible. No more — why it used to work earlier, but now it doesn’t?</li></ul><p>We also store these Docker images in our registry, so rolling back is as simple as redeploying a previous container version.</p><h3>Strategy Ops: Health Checks &amp; Real-Time Data</h3><p>Once your strategy is deployed, our <strong>Strategy Ops service</strong> keeps monitoring it:</p><ol><li><strong>Health Monitoring</strong>: We track CPU, memory usage, and process health. If your code crashes or spikes unexpectedly, you’ll get alerts.</li><li><strong>Live Quotes Integration</strong>: Each container can subscribe to the symbols it needs via our aggregator service, which channels real-time data over websockets.</li><li><strong>Scheduling</strong>: If your strategy is time-based (like end-of-day or end-of-month rebalancing), it runs exactly when you specify. We handle the scheduling under the hood.</li></ol><p>All of this is designed to help you worry less about ops and more about how to optimize your trades.</p><h3>Security and Whitelisting Packages</h3><p>We do limit which libraries can be installed, mainly to protect the environment. You’ll have access to the <strong>Wizzer SDK(s)</strong> and a set of vetted packages — includes all the libraries generally used by quants for data science &amp; algo trading ex., pandas, numpy, etc.</p><p>This restriction ensures:</p><ul><li><strong>No Malicious Code</strong>: Rogue packages can’t compromise the platform.</li><li><strong>Stable Environments</strong>: We keep everything updated, so you’re not stuck dealing with version mismatches or broken dependencies.</li></ul><p>If there’s a library you really need, we’re open to considering it. But by default, we keep the environment lean and secure.</p><h3>Why This Matters for Traders and Quants</h3><p>For many traders, the hardest part isn’t coming up with a strategy — it’s getting it up and running in a stable production environment without losing time or sanity. Our setup does the heavy lifting:</p><ul><li><strong>No More “Will It Run Here?”</strong>: If it works in your local environment with the same Docker image, it’ll work when deployed.</li><li><strong>No Manual Copy-Paste</strong>: Commit your code, watch the pipeline do its thing, then trade.</li><li><strong>Automated backtests</strong>: Commit your code, and our pipelines will run the backtests for you as a part of deployment process.</li><li><strong>Realtime-backtests: </strong>Run the backtests for your strategies in realtime with a single click — Backtests, and see the results in realtime.</li><li><strong>Full Accountability</strong>: Everything from code changes to container logs is tracked and auditable.</li></ul><p>We built this system because we believe quant/algo trading should be <strong>creative, not cumbersome</strong>. By blending a VSCode-like editor, Git-based version control, and container infrastructure, we’re bridging the gap between how software developers work and how traders want to work.</p><p>If you have questions about how these pipelines might fit your specific needs, feel free to reach out. We’re always happy to chat about how we can make trading workflows easier, and safer.</p><p>Thanks for reading — and stay tuned for more updates in our Wizzer Engineering Vision series.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7386a1705fcf" width="1" height="1" alt=""><hr><p><a href="https://medium.com/wizzercounsel/building-deploying-strategies-wizzers-code-editor-container-infrastructure-7386a1705fcf">Building &amp; Deploying Strategies: Wizzer’s Code Editor &amp; Container Infrastructure</a> was originally published in <a href="https://medium.com/wizzercounsel">Wizzer</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Inside Wizzer’s Data Pipelines & Workflows: How We Handle Market Data & Real-Time Streams]]></title>
            <link>https://medium.com/wizzercounsel/inside-wizzers-data-pipelines-workflows-how-we-handle-market-data-real-time-streams-9ea5ca7114a8?source=rss-21cbaa373d88------2</link>
            <guid isPermaLink="false">https://medium.com/p/9ea5ca7114a8</guid>
            <category><![CDATA[wizzer-tech]]></category>
            <dc:creator><![CDATA[Pawan Wagh]]></dc:creator>
            <pubDate>Thu, 20 Mar 2025 05:45:08 GMT</pubDate>
            <atom:updated>2025-03-20T05:45:08.573Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3A4hLifrZ6bG-57IoeFerw.png" /></figure><blockquote>In this part of our Wizzer Engineering Vision series, we’ll take a closer look at <strong>how data gets in &amp; stays accurate </strong>at Wizzer. In capital markets, <strong>data</strong> is everything — real-time prices, corporate actions, performance metrics, and more. I want to take you behind the scenes at <strong>Wizzer</strong> and share exactly what we’ve <strong>already built</strong> to manage market data. No vaporware or roadmaps — just the real stuff running in production right now.</blockquote><h3>Why Data Pipelines Matter ?</h3><p>Markets move fast, every exchange have their own formats for same data and even something as “simple” as a stock split can throw off your entire dataset if you don’t handle it carefully. At Wizzer, we quickly learned that <strong>if your data is messy, your trades might be too</strong> — so reliable pipelines aren’t a “nice to have,” they’re essential.</p><h3>Data Pipelines &amp; Workflows</h3><p>These workflows are orchestrated by either <strong>Airflow</strong> or <strong>custom cron jobs</strong> (depending on complexity and scheduling). A typical job might run around market close or in the early hours of the morning:</p><ol><li><strong>Fetch</strong>: Most of the workflows are polled based. They query/download data from multiple sources. ex., “Daily Instruments Master” workflow which downloads the master files from exchanges, transforms them into a standard format, feeds this data into an indexing engine which then gets converted into an index which can be stored on frontend for offline &amp; blazingly fast instruments search.</li><li><strong>Validate</strong>: A quick sanity check — no bizarre or missing fields, no incorrect date stamps. ex., Whether the closing prices for a sample of instruments is same for the last 7 days.</li><li><strong>Store</strong>: We then place the validated data into our storage layer, where it’s available for other workflows to consume.</li></ol><p>It’s not glamorous, but it’s crucial. Missing a single day’s closing prices can skew your entire backtesting or performance analysis. Automating this routine saves us from frantic 11 p.m. Slack messages about “Why is yesterday’s data missing?”</p><h3>Real-Time Data with NATS.io &amp; Our Aggregator Service</h3><p>While daily files are critical for end-of-day and historical analysis, <strong>real-time</strong> streaming is where the action happens. At Wizzer, we’ve set up a pipeline to connect directly to the exchange for <strong>live data</strong>, and we use <strong>NATS.io</strong> to streamline it internally.</p><p>Here’s a quick breakdown:</p><ol><li><strong>Exchange Connectivity</strong>: We have a component that receives real-time prices from the exchange.</li><li><strong>NATS.io</strong>: This data is published to NATS, a high-performance messaging system that handles distributing these market updates to other services.</li><li><strong>Aggregator Service (Go + Redis)</strong>: A custom service, written in Golang, subscribes to relevant topics on NATS. It keeps track of <strong>who</strong> (which client or strategy) needs which instrument’s data. We store those “subscriptions” in Redis for quick lookups.</li><li><strong>WebSocket Streams</strong>: Finally, our aggregator streams the relevant quotes to each subscribed client or strategy over a WebSocket connection.</li></ol><p>Why go through all these hoops instead of a single direct feed? Because we have to manage <strong>hundreds — sometimes thousands — of simultaneous subscriptions</strong>, each with different latency requirements, symbols, and user privileges. Redis helps us quickly match “Client X” to “Symbol Y,” while NATS keeps the messaging layer fast and efficient.</p><h3>Data Storage: MongoDB &amp; ClickHouse</h3><p>Once data (real-time, daily, or historical) arrives, it’s <strong>saved</strong> in two main places:</p><ul><li><strong>MongoDB</strong>: Often used for operational data or ephemeral states, like tracking user sessions or short-term data points that need quick writes and flexible schemas.</li><li><strong>ClickHouse</strong>: Our go-to for heavy analytical queries, historical candles, and <strong>backtesting</strong> data. If you’re storing millions of rows of price data, ClickHouse performs well for analytics and aggregation.</li></ul><h3>Incremental Backups &amp; AWS S3</h3><p>Since losing historical data would be catastrophic, we have <strong>incremental backups</strong> pushed to <strong>AWS S3 buckets</strong>, with both versioning and replication turned on. That way, even if a pipeline job goes haywire and overwrites something, we can roll back to a stable version. Storing these snapshots externally also protects us against local hardware failures or unexpected data corruption.</p><h3>Daily Performance Analytics</h3><p>Wizzer also tracks <strong>daily performance</strong> snapshots, focusing on:</p><ul><li><strong>Strategy Capital Allocation</strong>: How much capital each strategy uses, helping us spot imbalances or risk exposures.</li><li><strong>Daily Performance Metrics</strong>: Gains/losses, volatility, drawdowns — anything that might influence a trader’s or advisor’s decisions.</li><li><strong>Asset-wise allocation &amp; snapshots: </strong>Capital, trades, margins, realised &amp; unrealised P&amp;L for each strategy across asset classes.</li><li><strong>Advices statistics: </strong>Statistics for advices given out by advisor’s, performance of advices received &amp; executed by clients, and reporting.</li></ul><p>All of these analytics &amp; a lot more land in <strong>ClickHouse</strong> too. Because it’s designed for columnar storage and fast queries, we can run quick aggregates or big, multi-day queries without punishing our overall system performance.</p><h3>Python SDK for Historical &amp; Real-Time Access</h3><p>For backtesting and strategy deployment, we offer a <strong>Python SDK</strong> that taps into both <strong>historical</strong> and <strong>live</strong> data. This lets users:</p><ul><li><strong>Pull candles</strong> from a certain date range to test a new momentum strategy.</li><li><strong>Stream live quotes</strong> to see if their stop-loss levels get triggered in real time.</li><li><strong>Access fundamentals</strong> if they’re building a more advanced model that factors in earnings or financial ratios.</li></ul><p>By standardizing how we deliver data through an SDK, we ensure each deployed strategy on Wizzer’s platform can call the same functions, rather than reinventing the wheel or hacking multiple APIs together.</p><h3>Airflow &amp; Custom Cron: A Blend of Scheduling</h3><p>While <strong>Airflow</strong> is our main workhorse for more complex, multi-step workflows (like a job that runs data validation, then corporate action adjustments, then sends out an alert), we still rely on <strong>classic cron jobs</strong> for simpler tasks. Some pipelines don’t need the overhead of a full DAG (Directed Acyclic Graph) or detailed scheduling UI. For instance:</p><ul><li><strong>Cron-based</strong> job to fetch the trades register from the backoffice &amp; send it to the workflow for performance analytics.</li><li><strong>Airflow DAG</strong> to orchestrate a multi-step pipeline that merges data, checks for anomalies, and triggers a Slack notification if something looks off.</li></ul><p>This mix gives us the <strong>best of both worlds</strong>: we don’t over-engineer simple tasks, but we have robust tools ready when a pipeline gets more intricate.</p><h3>Putting It All Together</h3><p>When you zoom out, here’s how the flow typically looks:</p><ol><li><strong>Daily Data</strong>: Airflow/cron for scheduled or batch jobs.</li><li><strong>Adjustments</strong>: Another job updates historical records for events like splits or dividends.</li><li><strong>Real-Time Stream</strong>: Live ticks are ingested via NATS.io, passed to our aggregator service in Go, which tracks subscriptions with Redis.</li><li><strong>Data Storage</strong>: MongoDB for operational/flexible data, ClickHouse for large-scale analytical tasks.</li><li><strong>Analytics</strong>: Performance snapshots, strategy allocations, and more stored in ClickHouse, where we can easily query them.</li><li><strong>Access Layer</strong>: A Python SDK that delivers both historical and real-time data to deployed strategies.</li></ol><p>All the while, <strong>incremental backups</strong> ensure we never lose data, and <strong>Airflow</strong> plus <strong>custom cron</strong> keep these workflows running like clockwork.</p><h3>Why This Matters</h3><p>So why bother laying all this out? If your historical records are off by even a slight margin, your backtests could look great on paper but fail in the real world. If your live feed lags or drops during high volatility, you might miss crucial signals — or worse, place orders on stale information.</p><p>By keeping these workflows straightforward, focusing on reliability, and storing backups offsite, we give traders and advisors the <strong>peace of mind</strong> they need to innovate. They don’t have to wonder if we missed a corporate action or if a daily data job failed silently. And that sense of reliability leads to better decisions, and hopefully better returns for everyone involved.</p><h3>Conclusion</h3><p>At Wizzer, we believe in <strong>building from the ground up</strong> — solving the daily data pains before promising moonshot features. Our data pipelines handle real-time quotes, historical data, financial events ex., corporate actions, and incremental backups with a blend of open-source tools (NATS.io, MongoDB, ClickHouse) and custom scripts (Go, Python SDK, Redis aggregator).</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9ea5ca7114a8" width="1" height="1" alt=""><hr><p><a href="https://medium.com/wizzercounsel/inside-wizzers-data-pipelines-workflows-how-we-handle-market-data-real-time-streams-9ea5ca7114a8">Inside Wizzer’s Data Pipelines &amp; Workflows: How We Handle Market Data &amp; Real-Time Streams</a> was originally published in <a href="https://medium.com/wizzercounsel">Wizzer</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Strategy subscription & automated execution]]></title>
            <link>https://medium.com/wizzercounsel/strategy-subscription-automated-execution-fc58118c0a34?source=rss-21cbaa373d88------2</link>
            <guid isPermaLink="false">https://medium.com/p/fc58118c0a34</guid>
            <category><![CDATA[wizzer-tech]]></category>
            <dc:creator><![CDATA[Pawan Wagh]]></dc:creator>
            <pubDate>Tue, 28 Jan 2025 06:00:48 GMT</pubDate>
            <atom:updated>2025-01-28T06:00:48.178Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5IP6d9b2zNKnpNOpETvGCg.png" /></figure><p>Ever wonder if Quant Trading Strategies could be just subscribed like Apps. Subscribe and they just turn on and start giving you their advices. They get upgrades — better backtesting tweaks conditions, market regime changes and something upgrades in the strategy and so on. You pay for the advice just like your pay for Apple Music app. If you don’t upgrade — you get the freemium version. If it’s good enough for you, go on help yourself. If you like to be on the cutting edge — keep it upgraded.</p><p>At <em>Wizzer</em>, we’ve designed a platform that addresses the entire lifecycle: from how advisors publish their strategies, to how clients subscribe (or unsubscribe) to those strategies, handle subscriptions that expire, and even how new customers get onboarded.</p><h3>Problem —</h3><p>Offering quant products/strategies can be complex. Advisors need a seamless &amp; simple way to publish strategies, while clients demand an magical experience to consume those products.</p><p><strong>User experience —</strong></p><ul><li><strong>Advisors</strong> need an easy way to publish &amp; maintain multiple versions ot the strategies.</li><li><strong>Clients</strong> (both existing and new) must seamlessly onboard, subscribe, and fund their accounts for timely execution.</li><li>Both, advisors &amp; clients, should be communicated with realtime updates.</li></ul><p><strong>Complex Lifecycle Management —</strong></p><p>Strategies go through continuous updates, versioning, and deployments. Meanwhile, clients subscribe, unsubscribe, and sometimes let subscriptions expire.</p><ol><li><strong>Operational Continuity:</strong><br>Strategies must run reliably, even if a user has not yet allocated capital. Capital allocation can happen later, but we still need the container (strategy) up and ready.</li><li><strong>Scalability &amp; Automation:</strong><br>As more clients subscribe, orchestration must happen automatically and on-demand. A fully manual approach leads to error and reduces trust in the platform.</li></ol><h3>Solution —</h3><p>Wizzer addresses these challenges by providing an integrated <strong>Strategy Execution Platform</strong> with the following capabilities:</p><ol><li><strong>Advisor‐Friendly Code Publishing</strong></li></ol><ul><li>Advisors push strategy updates to a Git‐based <strong>Strategy Library</strong>.</li><li>A <strong>Continuous Deployment</strong> pipeline automatically builds container images and stores them in an <strong>Image Registry</strong>.</li></ul><p><strong>2. Seamless Client Subscription &amp; Onboarding</strong></p><ul><li>The moment a client clicks “Subscribe,” the latest container image is fetched and deployed — independent of capital allocation.</li><li>Notifications ensure clients know they’ve successfully subscribed.</li></ul><p>3. <strong>Automated Lifecycle Management</strong></p><ul><li><strong>Unsubscribing</strong> or <strong>subscription expiry</strong> triggers a container teardown, plus notifications for “Strategy uninstalled” and “Product unsubscribed.”</li><li><strong>Event‐driven architecture</strong> means each major action (subscribe, unsubscribe, expire) also raises an event that other systems can respond to (e.g., notifications, billing updates).</li></ul><p><strong>4. New Client Onboarding Flow</strong></p><ul><li>If someone is a visitor — who wants to try out the product, they’re guided through account creation, risk survey, payment, and brokerage account setup. Once that’s done, they can add capital and let the strategy run.</li></ul><p><strong>5. Real-time Notifications</strong></p><ul><li>Advisors receive build confirmations.</li><li>Clients get messages (via <strong>Email/SMS/WhatsApp</strong>) whenever a strategy is deployed, deactivated, or unsubscribed.</li></ul><p>This solution ensures:</p><ul><li><strong>Seamless user experience</strong>: Clients and advisors see a straightforward, guided process.</li><li><strong>Scalability</strong>: Systems automatically handle new or updated strategies, new user onboarding, and mass unsubscribes without manual intervention.</li><li><strong>Visibility</strong>: Clear, timely notifications at each step reinforce user confidence.</li></ul><h3><strong>Engineering —</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kcV2-UFzmGk47x7PegMQug.png" /><figcaption>Fig. Architecture &amp; component blocks</figcaption></figure><h3>Architecture Components</h3><ol><li><strong>Wizzer Backend</strong></li></ol><p>The central orchestration engine for the entire platform. This is where all the integration with client account, trading apis, exchange communication, and execution happens.</p><p><strong>2. Strategy Library (Git Server)</strong></p><p>A version-controlled repository where strategy code is stored. It is essentially a <strong>Git server</strong> and acts as the storage for strategy code just like GitHub or Bitbucket.</p><p>Advisors push updates to strategy code here. Each push creates a new version or branch. The Git server tracks changes and commits, ensuring strategies are versioned &amp; auditable.</p><p><strong>3. Continuous Deployment Pipeline</strong></p><p>Automates the process of building, testing, and deploying strategy container images. When a new version is pushed to the <strong>Strategy Library</strong>, this pipeline is triggered, and it builds the strategy into a <strong>container image</strong> (e.g., Docker), ensuring that the strategy includes all required dependencies and SDKs — Data Pipelines SDK, Wizzer SDK, pandas, etc — for execution.</p><p>Once successful, the image is <strong>tagged</strong> and <strong>stored</strong> in an image registry for easy retrieval by the deployment services.</p><p>4. <strong>Image Registry</strong></p><p>A repository for container images that stores the latest versions of the containerized strategy.</p><p><strong>5. Strategy Execution Orchestrator</strong></p><p>The heart of this entire execution platform responsible for the deployment and lifecycle of strategy containers. Once the client subscribes to a product, the lastest/or subscribed version of the strategy image will be pulled by this orchestrator for deployment.</p><p>While deployment, this component ensures that the deployed container is initialised &amp; tagged with user metadata — Account Id, Permissions, access tokens, etc — by injecting via environment variables. This ensures, when a strategy publishes events/actions, then underlying wizzer SDK will auto inject these metadata into strategy events/actions which then gets passed to the Wizzer backend for execution.</p><p>It monitors the health and status of the strategy containers and may take corrective actions (e.g., redeploy if needed).</p><p><strong>6. Strategy — Container Execution Platform</strong></p><p>The runtime environment for the strategy containers, where the strategy is running. This platform will take care of all the resource allocation — CPU, Memory, Storage — to ensure strategies are executing seamlessly.</p><p>Another important role of this component is ensuring the strategies are running within their own isolated, independent, and sandbox environments. This guarantees no multiple versions of the same or different strategies are conflicting or posing any risk to each other.</p><p><strong>7. Realtime Quotes Service</strong></p><p>Responsible for live market data that strategy can subscribe to for real-time trading decisions.</p><p><strong>8. Data Warehouse</strong></p><p>Centralized storage for historical and fundamental data. Data can be fetched as needed by the strategies<strong> </strong>on-demand using Data Pipelines SDK. This SDK will be pre-packaged with defined apis for fetching data.</p><p><strong>9. Events Broker</strong></p><p>Manages event-driven communication across different components of the system, and decouples different system components by handling events asynchronously.</p><p>Strategy actions occur — a client subscribes, or a stop-loss is hit in a strategy — an event is triggered and passed to the <strong>Events Broker</strong>. The broker distributes these events to relevant <strong>Event Consumers</strong>. It ensures scalability by handling high throughput of events and can trigger further workflows or updates (such as notifications or strategy adjustments).</p><p><strong>10. Events Consumers</strong></p><p>Process events received from the <strong>Events Broker</strong> and trigger relevant workflows &amp; call REST apis.</p><p><strong>11. Notification Channels (Email/SMS/WhatsApp)</strong></p><p>Real-time communications to clients via their preferred mediums.</p><p>Below are few keys user journeys expressed as sequence diagrams which describes various events/actions a client or strategy will go through.</p><ol><li><strong>Advisor publishes code</strong></li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2e92SyMpAbS_gbNFeqe8dQ.png" /><figcaption><strong>Advisor publishes code</strong></figcaption></figure><p><strong>Key Points:</strong></p><ul><li>Every code commit triggers the pipeline to produce a new container image.</li><li>Advisors are kept in the loop via email/SMS/WhatsApp.</li></ul><p>2. <strong>Client with no Wizzer account</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-h_t_RKNsXhTfc1gQNd63Q.png" /><figcaption>Client with no Wizzer account</figcaption></figure><p><strong>Key Points:</strong></p><ul><li>Combines user registration, risk management, payment, and broker onboarding into one guided flow.</li><li>Only after the user’s trading account is <strong>funded</strong> and capital is allocated does the strategy actively trade.</li></ul><p>3. <strong>Existing client subscribes/resubscribes to a product</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SdBk_zW0KTEygNlfcloaPQ.png" /><figcaption><strong>Existing client subscribes/resubscribes to a product</strong></figcaption></figure><p><strong>Key Points:</strong></p><ul><li>The client sees an immediate subscription success.</li><li>Prompt for capital allocation.</li></ul><p>4. <strong>Client unsubscribes from a product</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OBtBY4vQcBpr374OJNXKOg.png" /><figcaption><strong>Client unsubscribes from a product</strong></figcaption></figure><p><strong>Key Points:</strong></p><ul><li>Two separate notifications: <em>Strategy deactivated</em> and <em>Product unsubscribed</em>.</li><li>This ensures the client knows the strategy is uninstalled and the subscription is ended.</li></ul><p>5. <strong>Client’s subscription expires</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CVdDYaapCvHqNrBuwK1-hQ.png" /><figcaption><strong>Client’s subscription expires</strong></figcaption></figure><p><strong>Key Points:</strong></p><ul><li><strong>Subscription Manager</strong> automatically checks for expiration.</li><li>Workflow orchestrates the unsubscribe process.</li></ul><p><strong>Wizzer</strong> orchestrates everything from code publishing to subscription management, capital allocation, and user onboarding. By separating the <strong>strategy deployment</strong> from <strong>capital allocation</strong>, we give clients the flexibility to decide when to inject funds — while still making sure the strategy environment is up and running.</p><p><strong>Highlights:</strong></p><ul><li><strong>Continuous Delivery</strong> ensures new strategy versions are always ready to deploy.</li><li><strong>Immediate Subscription</strong> flow means minimal wait time; capital can be added later.</li><li><strong>Notifications</strong> play a central role in user communication, especially during on/off boarding events or key changes (e.g., unsubscribing, expiring).</li><li><strong>Full Onboarding</strong> flow streamlines the setup for new clients, integrating KYC, payment, risk assessment, and broker account setup into one smooth experience.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fc58118c0a34" width="1" height="1" alt=""><hr><p><a href="https://medium.com/wizzercounsel/strategy-subscription-automated-execution-fc58118c0a34">Strategy subscription &amp; automated execution</a> was originally published in <a href="https://medium.com/wizzercounsel">Wizzer</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>