<?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 Maxim Piessen on Medium]]></title>
        <description><![CDATA[Stories by Maxim Piessen on Medium]]></description>
        <link>https://medium.com/@maximpiessen?source=rss-cb03c065269e------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*O7MtHJ_4EV5ugtf_lXHhsA.jpeg</url>
            <title>Stories by Maxim Piessen on Medium</title>
            <link>https://medium.com/@maximpiessen?source=rss-cb03c065269e------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 13:51:10 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@maximpiessen/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[From matrices to finished work: How AI runs Credix]]></title>
            <link>https://medium.com/credix/from-matrices-to-finished-work-how-ai-runs-credix-607b55570f5f?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/607b55570f5f</guid>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Mon, 13 Apr 2026 10:26:26 GMT</pubDate>
            <atom:updated>2026-04-14T12:13:01.600Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y61lnkBRJJTm0trtvnBLuA.png" /></figure><p>In 2012, I was a physics freshman at the University of Antwerp, grinding through 26 hours of math per week. I fell in love with matrices: elegant, powerful, foundational. Back then, they were tools for solving physics problems. Today, the same linear algebra powers the AI systems running our company.</p><p>The math hasn’t changed. What changed is what we do with it. At Credix, AI doesn’t assist our team. It performs finished work. Decisions and actions previously performed only by humans.</p><p><strong>Where we are today</strong></p><p>Every person at Credix builds their own data-driven reports, without writing a single line of SQL. Analysis that used to take days now takes minutes. This is possible because we invested in our data platform and built a unified semantic layer in Hex: a collection of curated, documented metrics that can be sliced across many dimensions. Every agent and employee at Credix works from the same source of truth.</p><p>75% of our code is written by AI. Some of our engineers have stopped writing code entirely, not because they’re redundant, but because they’ve moved up the stack to architecture, system design, and product thinking.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*b2_qi8o3iIyDKulA9fLydw.png" /><figcaption>Source — <a href="https://workweave.dev/">https://workweave.dev/</a></figcaption></figure><p>70% of our customer success cases are fully resolved by our AI agent, Aira. Not just simple FAQs. Aira handles complex debt renegotiations end-to-end. Finished work at scale, whilst beating industry CSAT benchmarks.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zdUNhaCyW35g4OF8ZEhvTQ.png" /><figcaption>Deflection and CSAT since March</figcaption></figure><p><strong>Where we’re going by the end of Q2</strong></p><p>Our platform will detect and fix its own issues before anyone notices. We’re building a system that monitors production 24/7, performs root-cause analysis, pulls in context from engineers when needed, and ships a code fix before a ticket is filed. By building a self-healing platform, our engineers can focus on what actually moves the business forward: shipping product, not firefighting incidents.</p><p>Our AI agents in collections will stop following pre-determined agent routing and start learning from outcomes. We set the goals, provide data, tools, and guardrails, and let the systems improve themselves, the way a junior analyst becomes a senior one, but faster. We expect these systems to outperform their human predecessors across support, fraud detection, and collections. This development will serve as the foundation so that any non-technical person at Credix can set up agents to drive finished work.</p><p>Our outbound sales will be agentic. One person orchestrating a system of many agents, across many verticals, at a scale no traditional team could match. The agents do the heavy lifting, and the person focuses on building relations and closing deals.</p><p><strong>Why this matters</strong></p><p>This isn’t automation for the sake of efficiency. It’s a fundamentally different operating model. A team of 50 performing like a team of 500. That leverage gives us the space to do deep thinking and stay focused on becoming Brazil’s market-leading B2B credit infrastructure.</p><p>You’ll hear from me again at the end of Q2. Expect proof points for everything above, and a forward-looking thesis on what comes next.</p><p>Max</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=607b55570f5f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/from-matrices-to-finished-work-how-ai-runs-credix-607b55570f5f">From matrices to finished work: How AI runs Credix</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[From idea to production without engineering: How we entirely switched to Lovable for frontend…]]></title>
            <link>https://medium.com/credix/from-idea-to-production-without-engineering-how-we-entirely-switched-to-lovable-for-frontend-20b569e0c8e6?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/20b569e0c8e6</guid>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Thu, 14 Aug 2025 06:39:12 GMT</pubDate>
            <atom:updated>2025-08-14T06:39:12.397Z</atom:updated>
            <content:encoded><![CDATA[<h3>From idea to production without engineering: How we entirely switched to Lovable for frontend development</h3><p>CrediPay is our B2B credit solution for the Brazilian market, powered by a set of scalable, developer-friendly APIs. Sellers access it through VTEX, direct API integration, or our Seller Platform. This web app showcases the full capabilities of our API suite, enabling our customers to start quickly and then gradually embed the functionality into their systems.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KBSPzds-UIXZ7Tr_yK7koA.png" /></figure><p>Until recently, the Seller Platform was fully managed by our engineering team. Every feature — from scoping to design, development, and deployment — was reviewed and refined by PMs, designers, and engineers. It worked, but it wasn’t the fastest way to move. We wanted a model where product managers could build, ship, and improve the platform without waiting in an engineering queue.</p><p>That is where Lovable came in. In just a few days, we rebuilt the entire Seller Platform using Lovable, matching all the existing functionality of the previous, engineer-managed version. And then we went far beyond. With a few iterations and prompts from our product managers, we drastically increased the platform’s feature completeness and functionality. Sellers now benefit from a richer, faster-evolving product, and PMs can act immediately on insights and feedback.</p><p>Of course, moving to a PM-led build process did not mean cutting corners on quality or security. The Lovable-powered platform follows the same best practices we’ve consistently enforced. Authentication and authorization flows run through JWT tokens with Auth0 integration, and token expiry is strictly enforced to protect sessions. Our backend endpoints continue to validate permissions and inputs before executing any request. In short, Lovable has given us speed and autonomy without compromising on the standards that matter.</p><p>Before the switch, the delivery process for any new feature looked like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9zvg379dKFRrDiFTdAjspg.png" /></figure><p>Now, it looks like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*m1NXLNE9otgNZHh7VwRzKw.png" /></figure><p>The flow is straightforward. A product manager defines and builds the feature directly in Lovable, iterating until it meets the product vision. From there, the feature moves through our new streamlined process, from prompt to staging to production, in a matter of hours rather than weeks.</p><p>That’s the big picture. Now, let’s lift the hood and see how it works:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KXW9UxF-RKBDY_LbTGv1lw.png" /></figure><p>Behind the scenes, Lovable syncs the generated code directly to GitHub, following our branching strategy. Merging to staging or production triggers a GitHub Action that injects environment variables, builds the Vite app, and pushes the build artifact (dist folder) to another repository (e.g., seller-platform-staging). Afterward, CloudFlare workers pick up the changes and serve the latest version on our custom domain. This setup enables a granular release process, allowing us to serve different app versions in multiple environments.</p><p>Since making the switch, the Seller Platform has transformed from a carefully engineered but slow-moving tool into a fast-evolving product that keeps pace with the needs of our sellers. Engineers can now focus entirely on complex backend challenges, while PMs take ownership of the customer-facing experience from start to finish.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aGtJYsO0XZyRhzT92uuYWQ.gif" /><figcaption>Performing a refund — A frontend feature that would take 2 weeks to develop, done in minutes</figcaption></figure><p>Seeing the impact on the Seller Platform, we are already thinking about where else we can apply the same principles. Our vision is simple: the person who best understands the problem should be able to deliver the solution — without unnecessary handoffs or delays.</p><p>If you want to see where this journey began, have a look at our earlier post: <a href="https://medium.com/credix/production-grade-vibe-coding-firebase-studio-bolt-lovable-and-replit-f10e0da320cc">Production-grade Vibe Coding: Firebase Studio, Bolt, Lovable, and Replit</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=20b569e0c8e6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/from-idea-to-production-without-engineering-how-we-entirely-switched-to-lovable-for-frontend-20b569e0c8e6">From idea to production without engineering: How we entirely switched to Lovable for frontend…</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Production-grade vibe coding — Firebase Studio, Bolt, Lovable, and Replit]]></title>
            <link>https://medium.com/credix/production-grade-vibe-coding-firebase-studio-bolt-lovable-and-replit-f10e0da320cc?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/f10e0da320cc</guid>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Thu, 01 May 2025 17:45:24 GMT</pubDate>
            <atom:updated>2025-05-01T17:45:24.292Z</atom:updated>
            <content:encoded><![CDATA[<h3>Production-grade vibe coding — Firebase Studio, Bolt, Lovable, and Replit</h3><p>2025 — the year anyone can build to-do apps, flashcard games, and travel-wishlist platforms. Over the past few months, vibe coding platforms have taken the internet by storm. We are all software engineers now. Yet, are we?</p><p>Let’s start by acknowledging a significant difference between your casual Saturday hobby project and managing a production-grade platform serving real customers. Key aspects include handling multiple environments, setting up deployment pipelines, rigorous testing, monitoring and alerting, authentication, secure integrations, and more.</p><p>As a fast-growing startup, we constantly ask ourselves two critical questions: How do we optimally allocate our human resources and efficiently use our capital? Engineers represent one of our most valuable assets, and optimizing their time and effort is key.</p><p>At Credix, we’re developing sophisticated credit infrastructure. We’ve also built and continue to maintain a front-end platform to allow non-integrated customers to interact with our infrastructure. Ideally, our engineers should concentrate solely on the most challenging problems, none of which arise from front-end tasks.</p><p>Can we envision a scenario where our engineers no longer write front-end code? Is it possible to empower our product managers through vibe-coding tools, without compromising our definition of “production-grade”?</p><p>Let’s find out!</p><h3>The test — outline</h3><p>We will put Firebase Studio, Bolt, Lovable, and Replit to the test, side by side. The test comprises three main goals:</p><ol><li><strong>Build a frontend application</strong> that showcases a paginated table.</li><li>Implement <strong>built-in user authentication</strong> with Auth0, our standard authentication provider across our technology stack.</li><li><strong>Fetch and visualize data via an API</strong>, leveraging the authenticated user’s JWT token.</li></ol><p>I asked our senior engineers how long they’d typically need to implement this from scratch. The consensus? It takes up to a month, especially when essential features such as client-side caching and robust pagination are included.</p><h3>The prompt</h3><p>AI actions rely on the quality of their inputs. With this in mind, I’ve crafted a detailed prompt using ChatGPT. This prompt includes clear guidelines, technical specifications for our Auth0 configuration and Credix API interface, and explicit expectations — much like a technical product spec you’d write before beginning development. Additionally, I’ve included a screenshot of our current platform to provide visual direction and context.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0p5FzrADXwzaam8NOxDTLQ.png" /></figure><p>For those interested, the full prompt can be found <a href="https://www.notion.so/credix/EXT-Prompt-1e6917cac58d80498769cc35ef193ec5?pvs=4">👉 here 👈</a>.</p><h3>The test —Who can build this?</h3><p>Each of the four platforms starts with the same initial prompt and screenshot. From there, we iterate solely via the chat interface to resolve bugs and errors. We’ll document the number of interactions each platform required to achieve a functional result and share notable highlights and challenges from the experience.</p><h4>☠️ <strong>Google Firebase Studio ( </strong>gave up after 14 interactions)</h4><ul><li>❌ After the initial prompt, the preview window did not open (bug?). Then, Firebase Studio was stuck in a loop trying to fix an error. It took me 7 prompts + copy/pasting network and console logs from the developer console to get it fixed. Then, it struggled getting auth0 working. In the end, it just skipped the Auth0 integration and mocked the table (thus not fetching data via API). The end.</li></ul><p>I firmly believe in Google’s unfair advantage over the long run, as they own the world’s data. Yet, when it comes to Firebase Studio, there’s still a lot of work to be done. Maybe they should acquire Lovable or Bolt and start from there?</p><h4>🤨 <strong>Replit (partial functionality — </strong>11 interactions)</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LGXt9-W6ZVXj7u5clnmgHQ.png" /></figure><ul><li>❌ Replit had difficulties setting up the Auth0 integration. To fix it, I had to copy and paste the network and console logs from the developer console.</li><li>❌ Instead of calling the Credix API, replit created its own endpoint, from which it serves mocked data. It thus failed completing the requested API integration.</li><li>✅ Pagination and caching properly implemented.</li><li>🙏 Some nice features: Before Replit starts coding, it builds a plan and suggests that you include more features. In my case, it asked if it had to build export functionality, table filtering, etc. Furthermore, it builds a visual preview before building the first app version. This allows you to see if it’s on the right track.</li></ul><h4>😲 <strong>Bolt (2!</strong> interactions)</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JN0kJB38RSqpSSO7S2C9IQ.png" /></figure><ul><li>❌ Just like Firebase Studio and Replit, Bolt had difficulties integrating Auth0. I had to search the logs/network again to point it in the right direction.</li><li>✅ It did build a workable version after giving it more information about the Auth0 bug.</li><li>✅ API order fetching, pagination and caching properly implemented.</li></ul><h4>🤯 Lovable (One-shot wonder)</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*flUMdoOVVdj3SHtZsPSFPw.png" /></figure><ul><li>🤯 I got a working app with one prompt. One. Prompt.</li></ul><h3>The winner — Lovable</h3><p>Before wrapping up the test and sharing the outcomes, I have to say I’m impressed by Lovable and Bolt. With just a handful of prompts, they delivered a solid, functional version — complete with a custom authentication flow and API-driven data fetching. Tasks that would typically cost our engineers weeks, knocked out in mere minutes. This sure is revolutionary.</p><h3>Ready to re-think our production stack — Yes!</h3><p>Lovable not only won first place in the test but also offers variable features that were on our checklist—features that we need to integrate a vibe-coding platform into our technology stack.</p><ul><li>Github sync—Lovable offers a two-way sync between its web platform and GitHub. I will write another post on how we’ll leverage this feature, but think about local debugging and setting up release/deployment pipelines.</li><li>In-app visual editing: You can select UI elements and edit basic CSS/HTML without having to prompt or code. This is a neat feature for our product managers, allowing them to iterate quickly on wording and small design changes.</li><li>Multiplayer — the name says it all. In a Figma-like way, we can collaborate on one project.</li></ul><p>That’s it for this post! Next up: <em>How did we integrate Lovable into our tech stack?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f10e0da320cc" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/production-grade-vibe-coding-firebase-studio-bolt-lovable-and-replit-f10e0da320cc">Production-grade vibe coding — Firebase Studio, Bolt, Lovable, and Replit</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Partnering with Taktile to drive velocity]]></title>
            <link>https://medium.com/credix/partnering-with-taktile-to-drive-velocity-6661c0e19e53?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/6661c0e19e53</guid>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Thu, 12 Sep 2024 08:42:17 GMT</pubDate>
            <atom:updated>2024-09-12T08:42:17.906Z</atom:updated>
            <content:encoded><![CDATA[<p>Recently, we partnered with Taktile, a decision-making platform for FinTechs. In this post, I want to outline why Credix, an engineering-first company, has chosen Taktile to outsource part of the technology stack.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sX-8Fm1KkjY3VT7AFAefCQ.png" /></figure><p>First things first, how would I explain Taktile in one paragraph?</p><blockquote>Taktile empowers both non-technical and technical users to create APIs for decision-making workflows. It integrates a wide range of data sources, both internal and external, and allows users to enhance these flows with Python scripts and custom machine learning models. The platform takes care of essential tasks like A/B testing, backtesting, deployments, unit and integration testing, and API endpoint generation. It provides you with the flexibility to learn and make adjustments on the fly, without the need for precious engineering time. In essence, it provides all the tools we need, allowing us to focus on more important aspects of the product rather than spending time on building non-differentiating infrastructure.</blockquote><h4>The decision</h4><p>Back to the decision-making: Should we outsource part of our infrastructure to Taktile or build it ourselves? We have built a small team of all-star players and want to keep the headcount low. Why? It forces us to only work on A-problems, creating solutions that drive real business value—solving challenges that bring us closer to our goal of becoming the market leader in B2B credit in Brazil. This strategy has proven successful: building, shipping, and capturing market share with our B2B BNPL product, CrediPay, in under ten months.</p><p>When doing roadmap planning, we quickly agreed that building the infrastructure to run decisions at scale is not our core business and will never be. At Credix, we put all our energy into building a proprietary dataset and contextual machine learning underwriting/fraud models that use this data. Everything else is noise. Hence, we decided to go into the market and look for a team that made it their sole mission to become the best at building decision-making infrastructure. We found Taktile.</p><p>Two things that impressed me about Taktile were the team and the product. Looking for a long-term technology partner is not justified by having superior technology—the longevity of a company and its products results from its people. We had an instant match with the co-founders Maik and Max, and we were impressed by how they thought about their business and the team they’ve assembled. Second, the Taktile platform has proven to be feature-complete for most of our (complex) use cases.</p><h4>Use-cases</h4><p>In the meantime, we use Taktile in several cases, ranging from highly complex credit underwriting flows and real-time fraud detection to buyer onboarding and automated credit operations. Let me explain the “automated credit operations” and “credit underwriting flows” to give you a grasp of the efficiency gains we get with Taktile.</p><p>We decided to challenge ourselves to get the first use case up and running in production in a few days. To start our experiment, we chose the “freeze buyers” use case. We disable the ability to consume more credit whenever buyers pay late. Until recently, we managed this through a retool dashboard, having a person look every day at the delinquent buyers and blocking those that were past due (luckily, the number of cases was not astronomical, given the quality of our underwriting models 😉). In an ideal world, we could run this check every hour, automatically freeze delinquent buyers, and have a manual review if we’re about to freeze a buyer with a high credit limit. Building those functionalities would take at least two weeks for two developers. In Taktile, a non-engineer built this flow in 2 days. In the flow, from different nodes, we read data from our BigQuery and Postgres databases, send notifications for manual review to Slack, and write the results for our APIs. This flow is now part of our core operations and has been running without any downtime or problems for the past few weeks.</p><p>A second use case is the migration of our underwriting flow. Our underwriting process involves analyzing vast amounts of transactional, market, and alternative data related to our buyers. Our engineering team focuses on developing proprietary systems to (i) source this data in innovative ways, (ii) store it efficiently at scale, and (iii) train models using these unique datasets. Once the data is gathered and analyzed, we leverage the insights to make critical underwriting decisions, such as assigning ratings, determining credit limits, and setting maximum payment terms. This is where Taktile plays a pivotal role. With Taktile, we’ve implemented a sophisticated decision-making framework that prioritizes cost-efficiency. We begin by applying “knockout rules” using the least expensive data sources before progressing to more costly options — whether in terms of direct data costs or computational resources. Taktile also enables us to create tailored underwriting flows for specific industries, company sizes, and other unique factors. Crucially, it provides full transparency in decision-making, allowing us to trace and refine our policies for each underwriting request, ensuring continuous learning and improvement.</p><h4>Looking ahead</h4><p>We’re excited to explore Taktile&#39;s latest features, integrate our machine-learning models into their workflows, develop more use cases, and strengthen our partnership with their team. By providing valuable feedback, we aim to help shape a category-defining product. As the Taktile platform evolves, we can concentrate even more on tackling the complex challenges of digital B2B credit in Brazil.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6661c0e19e53" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/partnering-with-taktile-to-drive-velocity-6661c0e19e53">Partnering with Taktile to drive velocity</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[CrediQuery: Harnessing AI to transform data processes at Credix]]></title>
            <link>https://medium.com/credix/crediquery-harnessing-ai-to-transform-data-processes-at-credix-a419196bf653?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/a419196bf653</guid>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Thu, 16 May 2024 16:13:36 GMT</pubDate>
            <atom:updated>2024-05-16T16:13:36.399Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zGpldjmRN7PobRiqY9anQA.png" /></figure><p><em>At Credix, we’re building the next-generation credit network. We take an engineering-first approach to succeed in this mission, embracing the newest technologies to bring efficiencies and new experiences. CrediQuery marks our latest achievement in these efforts. This AI-driven tool is revolutionizing how we gather insights from our ever-growing internal data lake.</em></p><h4>Transforming Data Interaction</h4><p>We strongly emphasize capturing and storing (transactional) data from the start. Having this data structured cleanly makes it perfect for getting business insights, such as “How much GMV did we finance last month?” “Which buyers didn’t transact for more than two weeks?” “What has been our revenue split between sellers?” etc. The only problem is that answering those questions requires SQL knowledge (a coding language to retrieve data from a database); a skill primarily present in our engineering team. Yet; the engineering team should focus on building the best product for our clients.</p><p>CrediQuery reduces the time required to answer such questions from 30 minutes to an astonishing two minutes. This slack-based chatbot is not just about speed; it’s about transforming our approach to data analytics. By allowing users to interact with our databases using natural language queries, CrediQuery fetches the required information, formulates the necessary SQL queries, and organizes the data into actionable insights in Excel format. It’s fast, efficient, and incredibly user-friendly.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UgHAuKriBbgaekv5m1x4AA.png" /></figure><h4>Learning and Adapting</h4><p>One of CrediQuery&#39;s most remarkable aspects is its ability to learn and adapt. Each interaction with the system allows for human feedback on the results. This feedback mechanism ensures that CrediQuery meets our current needs and evolves with our business, continuously enhancing its precision.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PCZ2siCIxv1XyF1lzGMf-A.png" /></figure><h4>The Future of AI at Credix</h4><p>Integrating AI into our business processes goes beyond operational efficiency. It involves a strategic vision of how these technologies affect our workforce and the industry. We are committed to pioneering AI integration that supports our team and offers them more opportunities to engage in meaningful and strategic work.</p><p>CrediQuery marks the beginning of a new era at Credix. We are working on a future where AI-driven tools like CrediQuery become integral to various facets of our internal and external operations and products, from risk assessment to customer service. We aim to keep pace with technological advancements and lead them, creating innovative solutions that drive our business forward and set new benchmarks for the industry.</p><blockquote>Are you interested in the internal workings of CrediPay? Feel free to reach out to maxim@credix.finance Happy to talk! Also, we’re hiring. Check out our <a href="https://www.linkedin.com/company/credix-finance/jobs/">open positions</a> or reach out directly if you want to contribute in building the next-generation credit network.</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a419196bf653" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/crediquery-harnessing-ai-to-transform-data-processes-at-credix-a419196bf653">CrediQuery: Harnessing AI to transform data processes at Credix</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Feature release: invest on Credix without the need for a web3 wallet]]></title>
            <link>https://medium.com/credix/feature-release-invest-on-credix-without-the-need-for-a-web3-wallet-bd0676ad2c9e?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/bd0676ad2c9e</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[development]]></category>
            <category><![CDATA[investing]]></category>
            <category><![CDATA[smart-contracts]]></category>
            <category><![CDATA[blockchain]]></category>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Fri, 01 Sep 2023 12:16:30 GMT</pubDate>
            <atom:updated>2023-09-01T17:20:53.518Z</atom:updated>
            <content:encoded><![CDATA[<h3>New feature release allows for investing without the need for a Web3 wallet</h3><p><strong><em>At Credix, we’ve recently rolled out smart contract-owned wallets (SCOW), making investing in our credit products as easy as sending USDC from your exchange of choice — no more need for web3 wallet solutions and custody infrastructures.</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YF7sOfB_eGYiAVphwYMO8g.png" /></figure><p>Our mission is to connect institutional investors with unique credit opportunities in Latin America, building on the most advanced technologies to achieve scale, and unlocking risk-adjusted returns through robust risk management and structuring capabilities.</p><p>Building on the most advanced technologies entails using blockchain technology to enable integrated accounting, settlement, and business logic — all in one atomic transaction. Hereby, we solve the reconciliation issue in financial services and eliminate manual interventions and errors.</p><p>The downside of building on blockchain technology is the need for a new kind of operational infrastructure to enable smart contract interactions. There’s a need for wallets to approve transactions and custody solutions to store tokens safely. For our investors, mainly large institutional funds and asset managers, this is a whole new paradigm. The majority is getting comfortable with the use of stablecoins and smart contracts for efficiency gains, yet setting up a wallet and custody infrastructure raises eyebrows and slows down the onboarding process.</p><h3>Introducing SCOW: smart contract-owned wallets</h3><p>You can think of a SCOW as a deposit address, with a hard-coded route for money flows. You can send funds from, e.g., Coinbase, Circle, Fireblocks, etc., to this address. Once the funds hit the SCOW, it automatically invests in the investment opportunity for which the SCOW was set up. Lastly, the SCOW also stores the exchange address of the investor, which is used to send repayments, so they settle on the investors’ exchange.</p><p>The below diagram shows the money and token flows for a SCOW investor (left) and a non-SCOW investor (right).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*k4BNF95qSheN_nuIc5S6lA.png" /></figure><p>A similar setup is being used for the counterparty to the transaction. Once a deal is fully funded, the principal is sent to the exchange of the borrowing party. For deal repayments, an on-chain collection account is set up to which the borrower can make repayments. All without the need for Web3 wallets.</p><blockquote><strong>‘’The investment process is reduced to sending USDC to a unique smart contract-owned wallet (SCOW) address. No need for custody of LP/tranche tokens. Repayments automatically settle in USDC on the exchange’’, says Maxim Piessen, CTO at Credix.</strong></blockquote><p>In summary, SCOW brings a variety of improvements to the transaction flow:</p><ul><li>No more need for web3 wallet solutions, and custody solutions; bringing a substantial UX improvement to investors and borrowers</li><li>USDC flows through our platform in a matter of minutes, shortening the time from fiat to fiat</li><li>SCOWs have a hard-coded route for money flows: capital can only go from the investor’s exchange account to the correct investment opportunity and back; as enforced by the smart contracts</li><li>As SCOWs are set up per transaction, per investor, we lower on-chain AML risks. Through our diligence partner, we receive the KYC/B confirmation from the investor before setting up the SCOW, and the capital originates from their exchange, eliminating the probability of commingling funds.</li></ul><p>We’re excited to keep pushing the boundaries of innovation as we improve the user experience of our stakeholders and support institutional investors in entering the space.</p><p>Stay up-to-date with our product releases by subscribing to our <a href="https://share-eu1.hsforms.com/1jAAAMPCdTpOl9Sx_K_7HrAfuste">newsletter</a> or following us on <a href="https://www.linkedin.com/company/credix-finance/">LinkedIn</a>. Inspired and interested in joining the team? Check out our open positions on our <a href="https://credix.notion.site/credix/Work-with-us-at-Credix-606913498d5a4f02949a641e738f9257">work at Credix page</a>.</p><p><em>The information contained herein is provided for marketing purposes only and should not be relied upon for investment decisions. This is not an offer to sell or a solicitation of an offer to buy securities.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bd0676ad2c9e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/feature-release-invest-on-credix-without-the-need-for-a-web3-wallet-bd0676ad2c9e">Feature release: invest on Credix without the need for a web3 wallet</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Unlocking New Possibilities: Tokenizing Existing Credit Facilities on the Credix Platform]]></title>
            <link>https://medium.com/credix/unlocking-new-possibilities-tokenizing-existing-credit-facilities-on-the-credix-platform-c8f548b977f9?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/c8f548b977f9</guid>
            <category><![CDATA[debt]]></category>
            <category><![CDATA[credit]]></category>
            <category><![CDATA[product]]></category>
            <category><![CDATA[fixed-income]]></category>
            <category><![CDATA[finance]]></category>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Thu, 18 May 2023 15:44:38 GMT</pubDate>
            <atom:updated>2023-05-18T15:44:38.635Z</atom:updated>
            <content:encoded><![CDATA[<blockquote>As a leading private credit portal connecting institutional investors with credit opportunities in Latin America, we at Credix are excited to announce our latest protocol upgrade, which now allows for the tokenization of existing credit facilities.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*S9kP4Ykzfk1JD58ekypjJA.png" /></figure><p>Previously, we were only able to bring new credit facilities on-chain, with the funding for these facilities happening on-chain as well. However, with the latest platform upgrade, funds like <a href="https://addem-capital.com/">Addem Capital</a> can now tokenize their existing credit deals, increasing the scalability of their business and tapping into a wider pool of potential investors.</p><h4>Here is how it works</h4><p>An asset originator or FinTech with an existing facility with Addem Capital, for example, can build up a track record of timely repayments. Our risk and structuring team takes up this historical payment data in the due diligence process to ensure the quality of deals brought to the platform.</p><p>A new structure for the deal is then created, which often involves a senior/junior tranche setup. Tokens are then issued to Addem Capital for both the senior and junior tranches, with no funds being moved in the process. From this point on, all repayments happen on the Credix platform, increasing the deal specific token value.</p><h4>Here is what it enables</h4><p>For fund managers like Addem Capital, the restructuring of deals from uni-tranche to senior/junior tranches allows them to sell part of their senior exposure to the liquidity pool or other investors on the Credix portal, freeing up capital to invest in other credit deals. They can keep the junior exposure, which serves as protection for the senior investor and offers them a higher yield than before.</p><p>The tokenization process also provides increased transparency, as investors can now track their investments in real time and have a clear view of the performance of the underlying credit facility. Additionally, this later on will (technically) allow for the tokenized credit facilities to be traded on secondary markets, providing increased liquidity for investors and allowing them to easily exit their positions.</p><p>This new upgrade provides tremendous efficiency gains and increased scalability for credit funds.</p><p>Stay tuned for more product updates and releases!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c8f548b977f9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/unlocking-new-possibilities-tokenizing-existing-credit-facilities-on-the-credix-platform-c8f548b977f9">Unlocking New Possibilities: Tokenizing Existing Credit Facilities on the Credix Platform</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Credix’s new monitoring feature unlocks the power of credit data]]></title>
            <link>https://medium.com/credix/credixs-new-monitoring-feature-unlocks-the-power-of-credit-data-78435d0d2030?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/78435d0d2030</guid>
            <category><![CDATA[finance]]></category>
            <category><![CDATA[debt]]></category>
            <category><![CDATA[credit]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[monitoring]]></category>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Tue, 11 Apr 2023 14:30:42 GMT</pubDate>
            <atom:updated>2023-04-11T14:30:42.246Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AEIzyo9UP4gBi-nShMqErQ.png" /></figure><blockquote>Credix has launched a new product feature that collects and standardizes all relevant information on the assets collateralizing the on-chain deals, providing investors with access to data on asset eligibility, portfolio covenants, collections, distributions, delinquencies, and defaults. The feature allows for efficient and accurate monitoring of investments, leading to more informed decision-making.</blockquote><p>Reliable and comprehensive data is critical to making informed investment decisions. At Credix, we understand this and have developed a system that provides the data our investors need to perform deep due diligence and monitor their investments.</p><p>Credit transactions consist of two main parts: the liability side and the asset side. On the liability side, we have smart contract logic that automatically sends interest and principal repayments to different investors based on the waterfall structure of the deal. More information <a href="https://medium.com/credix/credix-v2-flexibility-to-handle-complexity-2471c50f7ea9">here</a>.</p><p>On the asset side, we require our FinTech partners to post cash-generating assets as collateral for the facility. Every asset has its own characteristics, and repayment schedule. We carefully select eligible assets that meet our criteria. We also have portfolio covenants in place that act as rules on the aggregate portfolio of assets. All of this generates thousands of data points that are critical to monitoring the health of the collateral pool and predicting the future performance of the deal.</p><p>To ensure investors have access to this important data, we’ve developed a system that collects and standardizes all relevant information. We worked with our FinTech partners, monitoring agents, and investors to identify all available data points and define a data standard for facility monitoring. Our tooling makes it easy for stakeholders to upload required data in our standard, and automated checks ensure the correctness of the data. Those checks range from validating the formatting of the input values to recalculating the business logic of specific ratios and cross-checking with the portfolio’s collection account.</p><p>Once the data passes our quality filter, investors can access it on the platform. This data includes information on asset eligibility, portfolio covenants, collections, distributions, delinquencies, and defaults. At Credix, we’re committed to providing our investors with the data they need to monitor their investments and make informed decisions. By standardizing and automating our data collection and reporting, we aim to make this process as efficient and accurate as possible.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1IfAhuBZZS-w3EPh3Mm-Hw.gif" /><figcaption>An example of the facility report page (fictional data)</figcaption></figure><p>We focus on continuously improving our platform and providing even more value to our investors. Our next steps involve deep data integrations with our FinTech partners, allowing for real-time updates to be fed directly into our platform. This will enable investors to access up-to-date information on their investments, leading to more efficient and informed decision-making. In addition, we plan to gain access to loan-level data to provide investors with greater transparency and insight into the underlying assets. Finally, we are exploring the use of statistics and AI to improve our internal due diligence process and asset selection, ultimately helping us identify potential risks and opportunities early on in the investment process.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=78435d0d2030" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/credixs-new-monitoring-feature-unlocks-the-power-of-credit-data-78435d0d2030">Credix’s new monitoring feature unlocks the power of credit data</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Credix V2 — flexibility to handle complexity]]></title>
            <link>https://medium.com/credix/credix-v2-flexibility-to-handle-complexity-2471c50f7ea9?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/2471c50f7ea9</guid>
            <category><![CDATA[platform]]></category>
            <category><![CDATA[announcements]]></category>
            <category><![CDATA[defi]]></category>
            <category><![CDATA[stablecoin-cryptocurrency]]></category>
            <category><![CDATA[crypto]]></category>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Tue, 16 Aug 2022 14:34:32 GMT</pubDate>
            <atom:updated>2022-08-17T15:22:03.018Z</atom:updated>
            <content:encoded><![CDATA[<h3>Credix V2 — flexibility to handle complexity</h3><p><em>Debt capital markets are complex, and so are the financial products driving them. At Credix, we mix proven TradFi- and DeFi concepts to build the future of global credit markets. Today, we’re proud to announce the V2 of the Credix platform, featuring tranching, and arbitrary repayment schedules to allow for more complex deals.</em></p><p>Back in January, we launched the V1. Its main functionalities could be summarized as follows:</p><ul><li>Investors investing in the liquidity pool of a market</li><li>Investors getting LP tokens as a proof of their investment in the pool</li><li>Deals (bullet loans) are being created for borrowers, tapping into the liquidity pool</li><li>Deal repayments lead to an increase in LP token price and thus a positive yield for the investors</li></ul><p>This version, accompanied by our aggressive go-to-market strategy and thorough credit underwriting processes, paved the way to where we stand today: <strong>17.5M</strong> <strong>USDC in live deals</strong> and <strong>0% defaults</strong></p><p>In order to support bigger credit facilities, and cater to different investors’ needs; we’ve worked hard on implementing arbitrary repayment schedules and deal tranching in a generalistic fashion. Generalistic? We didn’t limit ourselves to known deal structures while designing and developing; <strong>we can easily support the most exotic schedule/tranche structures. Welcome to Credix V2!</strong></p><h3>Arbitrary repayment schedules</h3><p>Until now, we only supported bullet repayments. Meaning that the borrower would repay interest-only over xx periods of time and repay the principal in full at the end of maturity.</p><p>In reality, there is a wide range of repayment schedules possible: bullet with the principal at the end of maturity, bullet with the principal and all interest at the end of maturity, balloon, French amortization, German amortization, amortization with a grace period, …</p><p>The type of schedule is often matched to the cash-generating profile of the underlying assets. A bullet loan might be a good choice for a farmer financing crop seeds; as the farmer can repay its principal when harvesting its crops. Amortization is a good choice for mortgages as interest and principal are repaid along the deal’s lifetime at a fixed sum per month.</p><p>When creating a deal on Credix, you can now select the repayment schedule of your choice. Next to the pre-configured two most popular ones (bullet, and amortization) the new version allows for arbitrary repayment schedules. You can thus generate the schedule (expected principal and interest per repayment period) in e.g. Excel and send it to our smart contract through the app.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hvcBpCEgTwqYVV_y1p0WQw.gif" /></figure><h3>Arbitrary tranches and waterfalls</h3><p>Investors come in different shapes and sizes. Some are willing to take more risks than others if they are rewarded with higher returns. The concept of tranching was born to cater to those different risk-reward profiles. Tranching allows splitting up a deal into different parts. If we split the deal into 3 such parts, we call them the senior tranche, the mezzanine tranche, and the junior tranche.</p><p>Let’s look at an example of a bullet loan of 1M with an annualized interest rate of 15% and a performance fee (= the fee taken on interest repayments) of 10%:</p><ul><li>The <strong>senior tranche</strong> is often the largest and is funded by the liquidity pool. Let’s say this tranche makes up 80% of the deal. The expected APY for this tranche is 10.13%. The invested capital is protected by the mezzanine and junior tranches’ principal, and yield, meaning that they lose capital first when defaults of the repayments happen. The senior tranche is thus the tranche with the lowest risk.</li><li>The <strong>mezzanine tranche </strong>is funded by the underwriters after they’ve done the due diligence on the deal. Let’s say this tranche makes up 15% of the deal. The expected APY for this tranche is 22.5%. The mezzanine tranche is still protected by the junior tranche’s principal and yield.</li><li>The borrower itself funds the<strong> junior tranche</strong> to stimulate good repayment behavior. Let’s say this tranche makes up 5% of the deal. It can expect all proceeds after the senior- and mezzanine tranches reach their expected returns. In this example, the junior tranche could expect an APY of 40.5%</li></ul><p>(Quick sanity check: (0.1013*0.8+0.225*0.15+0.405*0.05) = 0.135 which is indeed the interest after the performance fee (0.15*0.9) )</p><p>Depending on the structure and profile of the deal, we can change the sizes and expected returns of the tranches. Furthermore, we can create single-tranche deals (senior only), or go up to 10 tranches.</p><p>Another challenge we’ve seen in the market is lockups of investments in more junior tranches. To circumvent this, we have built an advanced feature that allows underwriters to withdraw (part of) their initial investment and/or yield along the deal’s lifecycle if the underlying collateral is behaving well.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vURjJHX8TRsNg9wuMqY18A.gif" /></figure><h3>Bringing it all together</h3><p>With the launch of the V2, Credix brings proven TradFi concepts on-chain to handle more diverse and complex credit deals. Investors are able to pick investments according to their risk appetite, and trust the smart contract to handle the repayment flows and tranche waterfalls in an automated and efficient way.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*X85I5RTLO0AWCqra1f1_Kw.gif" /></figure><blockquote>Interested in learning more about Credix? <a href="https://docs.credix.finance/">Check our documentation</a><br>Interested in joining as an investor or borrower? <a href="https://app.credix.finance/">Start your journey</a></blockquote><p>Over the coming weeks we’ll share more details on the first deals going live on our V2. Special thanks for all the input to our underwriters, LPs and fintechs!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2471c50f7ea9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/credix-v2-flexibility-to-handle-complexity-2471c50f7ea9">Credix V2 — flexibility to handle complexity</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Credix’ layered security model]]></title>
            <link>https://medium.com/credix/credix-layered-security-model-bf357544031b?source=rss-cb03c065269e------2</link>
            <guid isPermaLink="false">https://medium.com/p/bf357544031b</guid>
            <category><![CDATA[wallet]]></category>
            <category><![CDATA[security]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[hacker]]></category>
            <category><![CDATA[solana-network]]></category>
            <dc:creator><![CDATA[Maxim Piessen]]></dc:creator>
            <pubDate>Wed, 03 Aug 2022 11:46:34 GMT</pubDate>
            <atom:updated>2022-08-03T11:46:34.389Z</atom:updated>
            <content:encoded><![CDATA[<p>At Credix, we’re building the future of global credit markets. We build on the most advanced decentralized technologies to achieve this goal, while always taking a security-first approach, both from a technology standpoint as from a development and methodology standpoint. That’s why we’ve introduced several layers of security to the Credix platform as visualized below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ksWVuP23gBJEnKDb5nbiNA.png" /></figure><h4>Tests — tests — tests</h4><p>We take testing seriously. This reflects itself in a testing suite covering unit tests in RUST to validate our Solana program code (smart contract), unit tests in javascript to ensure that our client (= abstraction layer to interact with the smart contracts) works as expected, and visual- and automated user-flow tests for our app to see that the application behaves the right way. Lastly, we’ve also built a fully-fledged digital twin in python. This digital twin mimics the behavior of the smart contract, allowing us to generate an infinite amount of end-to-end scenarios, which are then fed to the smart contract tests.</p><p>The test suite runs automatically on every change of the smart contracts, client, and app; ensuring that old features still function as expected when introducing changes.</p><h4>Security audits</h4><p>Upon the release of bigger features (e.g. tranching, secondary market, etc.), we work together with specialized smart contract audit firms. We leveraged Certik’s experience for the first audit back in December 2021. The report can be found <a href="https://drive.google.com/file/d/1Ys3OkMQZEfRGN2VWefLzqgZ0ZVTrA_gi/view?usp=sharing">👉 here 👈</a>. In June 2022, we had our V2 audited by Halborn. The report can be found <a href="https://drive.google.com/file/d/1duZ9k4RE5QQhLrYHjcwZL-OkdhzLBAuD/view?usp=sharing">👉 here 👈</a>.</p><h4>Stakeholder whitelisting</h4><p>All stakeholders have to be whitelisted by Credix. On a technical note, we issue a Credix pass on-chain after all legal, compliance and commercial obligations of our stakeholders are met. This Credix pass lists the permissions for the stakeholder; a borrower is only allowed to interact with the borrower-related program instructions, just like an investor is only able to invest and withdraw (after a specific lockup time). Every instruction does a check for this Credix pass, shielding the program from any external actors or attackers.</p><h4>KYC/KYB’d stakeholders only</h4><p>Credix is a fully permissioned marketplace; meaning that all stakeholders have to identify themselves. Individuals (e.g. accredited investors) have to go through a KYC (know your customer) process; businesses (e.g. institutional investors) have to go through a KYB (know your business) process. This process only needs to be performed once during onboarding. Once the onboarding is completed; an on-chain proof is issued to the wallet of the individual/business. This proof is needed to interact with the Credix platform.</p><h4>Non-transferable tokens</h4><p>The Credix platform leverages several tokens to keep track of the investments and accrued yield. The LP token is issued when investing in the liquidity pool of a market. Tranche tokens are issued when investing in junior tranches of specific deals.</p><p>Due to regulatory ambiguity, today none of our tokens can be made transferable as a KYB’d/whitelisted investor could transfer its LP tokens to a non-KYB/C’d investor. Therefore, all of our tokens are made non-transferable. This not only makes Credix comply with regulations but also stops hackers from transferring tokens from your wallet in case of a wallet exploit. We are working on an OTC desk and secondary market to make transfers between trusted stakeholders possible.</p><p>If you want to learn more about how we tackle security at Credix, I recently did an AMA with Halborn explaining the above and more. Check it out!</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F8ZLKCudPY7k%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D8ZLKCudPY7k&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F8ZLKCudPY7k%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/9dc1b9bdba99675a36a766d6a7c41444/href">https://medium.com/media/9dc1b9bdba99675a36a766d6a7c41444/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bf357544031b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/credix/credix-layered-security-model-bf357544031b">Credix’ layered security model</a> was originally published in <a href="https://medium.com/credix">Credix</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>