Lab49

🔄 Yesterday's FDC3 for the Web webinar hosted by interop.io was great! Big thanks to Leslie Spiro and Julianna Langston for hosting. The ease and simplicity of using FDC3 thanks to the new release from the team is going to make a material impact on the adoption of FDC3 as a standard. 👀 Don't forget to have a peek at the GitHub repository in the FINOS org: https://lnkd.in/enANq7TJ Key takeaways from the webinar: 🔹 FDC3 integration is now incredibly simple, just 'await getAgent();' 🔹 Browser-based desktop agents eliminate need for desktop installations 🔹 Full TypeScript support which gives you type safety in your code, but also improves developer experience in IDEs 🔹 Seamless integration across desktop and browser environments is an active area of development, the team is currently working on Agent Bridging To borrow Leslie's metaphor, FDC3 for the Web acts as a universal key, enabling apps to work across any (or no!) container environment. Additionally, AI-driven workflows are emerging which leverage FDC3 to integrate your data with AI agents. Cutting edge stuff! 🎪 Lab49 will be attending a few upcoming events like the interop.io Client Forum on April 29th, and the Open Source in Finance Forum in New York in October. See you there! #FinTech #OpenSource #FDC3 #Innovation #DesktopInterop

Last week Brian Wm. McAllister and I got to attend Future of Portfolio Management & Artificial Intelligence at Columbia University Mathematics of Finance (MAFN) hosted by Rebellion Research. Big thanks to Alexander Fleiss! 🚀 From Data to Dominance: What’s Next for AI in Investment Management? One thing was clear: AI is no longer an edge case—it is the core infrastructure for the future of investing. 🔹 Alpha is being found in narrative signatures, not just price signals. 🔹 LLMs and agentic AI are enabling new mid-frequency and systematic strategies. 🔹 But none of it works without clean, contextual, and explainable data. 🔹 The next-gen trading stack requires AI-ready data pipelines, resilient ML models, and deep domain alignment. At Lab49, we’re helping the world’s top financial institutions turn data complexity into AI-enabled opportunity. 📊 From digital twins to dynamic workflows, the future is being built now—and it’s intelligent by design. #AI #ML #Fintech #CapitalMarkets #DataEngineering #InvestmentManagement #Lab49 #Quant #LLM #AgenticAI

Joseph McSorley, mobile team lead at Lab49, presents a lightweight SwiftUI navigation framework that simplifies programmatic push navigation, as well as modal, and alert presentation. It also provides lifecycle management for observables (ViewModels in MVVM) associated with each screen in your app.    Here’s part one of the three part the series: https://lnkd.in/edkHHYaa   Check out the following Github links for more details:   - iOS 16: https://lnkd.in/eUPJq_aU - iOS 17+: https://lnkd.in/eEBra5ic   #swift #swiftui #ios #navigation

Our own Enmanuel Rivera just published a new tech talk!   Headless design systems are a great way to abstract branding out of style authoring. This helps shift responsibility of brand synchronization to product and design teams.    Figma variables are a great way to define design tokens that keep wireframes consistent. Variables are also used to produce build inputs that propagate branding across a applications without touching code.    Check out the new tech talk video right here: https://lnkd.in/eyhhSyhq   #designsystems #figma #collaboration #shiftleft

Building team culture is incredibly hard. It takes a team to do it, and it’s more than simply writing down your values. Having fun together is important: it’s what makes for a well-rounded shared experience.   So at Lab49, we do some pretty fun stuff! In February, the Interactive Practice ran a global contest to see who could get the highest score on https://cantunsee.space!   Because we have multiple locations, we ran multiple events around the world to get the team together and fight it out to see who would come out on top.    After a full week of events and contest entries, here’s the results:   🥇 Lavi Gupta with a score of 7,930 🥈 Joseph McSorley, also with a score of 7,930, but Lavi submitted his entry first! 🥉 Tied for third with scores of 7,730: Anshul Khare and Máté Végh   Congrats to all the winners!   Stay tuned here, many more contests to come at Lab49 in the future 🚀🎉🏆   #teambuilding #events #contest #lab49   (Sidenote, thanks to Alex Kotliarskyi for creating Can’t Unsee!)

𝗞𝗲𝘆 𝗧𝗮𝗸𝗲𝗮𝘄𝗮𝘆𝘀 𝗳𝗿𝗼𝗺 𝗔𝗜 𝗘𝘃𝗲𝗿𝘆𝘁𝗵𝗶𝗻𝗴 𝗚𝗹𝗼𝗯𝗮𝗹 𝗖𝗼𝗻𝗳𝗲𝗿𝗲𝗻𝗰𝗲 – 𝗔𝗜 𝗶𝗻 𝗙𝗶𝗻𝗮𝗻𝗰𝗶𝗮𝗹 𝗦𝗲𝗿𝘃𝗶𝗰𝗲𝘀   This week, I was at AI Everything.  Here are some key takeaways from one of the most interesting Panels - “Scaling AI or scaling risk?”   𝟭. 𝗧𝗼𝗽 𝗔𝗜 𝗨𝘀𝗲 𝗖𝗮𝘀𝗲𝘀 𝗗𝗿𝗶𝘃𝗶𝗻𝗴 𝗜𝗺𝗽𝗮𝗰𝘁 🔹 𝗖𝘂𝘀𝘁𝗼𝗺𝗲𝗿 𝗘𝗻𝗴𝗮𝗴𝗲𝗺𝗲𝗻𝘁 & 𝗢𝗽𝗲𝗿𝗮𝘁𝗶𝗼𝗻𝘀: Enhancing digital channels, frontline tools, and customer service 🔹 𝗥𝗶𝘀𝗸 & 𝗖𝗼𝗺𝗽𝗹𝗶𝗮𝗻𝗰𝗲: Fraud detection, credit decisioning, and monitoring trading communications 🔹 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜: Boosting developer productivity (30-40K coders at Citi use Gen AI!) and powering chatbots/wealth management tools 🔹 𝗢𝗽𝗲𝗿𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆: Optimizing cash centers, reducing client churn, and automating workflows   𝘒𝘦𝘺 𝘵𝘢𝘬𝘦𝘢𝘸𝘢𝘺: 𝘶𝘴𝘦𝘴 𝘤𝘢𝘴𝘦𝘴 𝘩𝘢𝘷𝘦𝘯'𝘵 𝘦𝘷𝘰𝘭𝘷𝘦𝘥 𝘮𝘶𝘤𝘩 𝘪𝘯 𝘵𝘩𝘦 𝘭𝘢𝘴𝘵 𝘺𝘦𝘢𝘳𝘴 𝘣𝘦𝘤𝘢𝘶𝘴𝘦 𝘣𝘢𝘯𝘬𝘴 𝘢𝘳𝘦 𝘢𝘷𝘦𝘳𝘴𝘦 𝘵𝘰 𝘳𝘪sk 𝟮. 𝗗𝗮𝘁𝗮: 𝗧𝗵𝗲 𝗗𝗼𝘂𝗯𝗹𝗲-𝗘𝗱𝗴𝗲𝗱 𝗦𝘄𝗼𝗿𝗱 ✅ 𝗜𝗻𝘁𝗲𝗿𝗻𝗮𝗹 𝗗𝗮𝘁𝗮 𝗙𝗶𝗿𝘀𝘁: Most banks underuse existing structured data (transactions, CRM) - “Stop making customers fill forms—leverage what you already have!” ⚠️ 𝗘𝘁𝗵𝗶𝗰𝗮𝗹 𝗚𝘂𝗮𝗿𝗱𝗿𝗮𝗶𝗹𝘀: Social media and hyper-personalized data (e.g., grocery spend) raise red flags. Consent and transparency are non-negotiable 🌐 𝗔𝗹𝘁𝗲𝗿𝗻𝗮𝘁𝗶𝘃𝗲 𝗗𝗮𝘁𝗮: Pilots with government/public data (income, property) show promise but require rigorous governance 𝘒𝘦𝘺 𝘵𝘢𝘬𝘦𝘢𝘸𝘢𝘺: 𝘌𝘹𝘵𝘳𝘢𝘤𝘵 𝘷𝘢𝘭𝘶𝘦 𝘧𝘳𝘰𝘮 𝘵𝘩𝘦 𝘦𝘹𝘪𝘴𝘵𝘪𝘯𝘨 𝘥𝘢𝘵𝘢 𝘣𝘦𝘧𝘰𝘳𝘦 𝘵𝘩𝘪𝘯𝘬𝘪𝘯𝘨 𝘰𝘧 𝘴𝘦𝘢𝘳𝘤𝘩𝘪𝘯𝘨 𝘧𝘰𝘳 𝘮𝘰𝘳𝘦 :) 𝟯. 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗔𝗜: 𝗛𝘆𝗽𝗲 𝘃𝘀. 𝗥𝗲𝗮𝗹𝗶𝘁𝘆 🤖 𝟮𝟬𝟮𝟱-𝟮𝟬𝟮𝟳: Expect semi-autonomous agents (e.g., data engineering assistants), but full autonomy is years away due to legacy infrastructure and regulations 🔧 𝗟𝗼𝘄-𝗛𝗮𝗻𝗴𝗶𝗻𝗴 𝗙𝗿𝘂𝗶𝘁: Start with non-critical tasks (e.g., summarizing emails, sales script analysis) to build trust 𝘒𝘦𝘺 𝘵𝘢𝘬𝘦𝘢𝘸𝘢𝘺: 𝘛𝘦𝘤𝘩𝘯𝘰𝘭𝘰𝘨𝘺 𝘮𝘰𝘷𝘦𝘴 𝘦𝘹𝘵𝘳𝘦𝘮𝘦𝘭𝘺 𝘧𝘢𝘴𝘵. 𝘛𝘦𝘤𝘩𝘯𝘰𝘭𝘰𝘨𝘪𝘴𝘵𝘴 𝘸𝘢𝘯𝘵 𝘵𝘰 𝘪𝘮𝘱𝘭𝘦𝘮𝘦𝘯𝘵 𝘣𝘶𝘵 𝘵𝘩𝘦 𝘣𝘶𝘴𝘪𝘯𝘦𝘴𝘴 𝘯𝘦𝘦𝘥𝘴 𝘵𝘪𝘮𝘦 𝘵𝘰 𝘣𝘦 𝘤𝘰𝘮𝘧𝘰𝘳𝘵𝘢𝘣𝘭𝘦 𝟰. 𝗥𝗲𝗴𝘂𝗹𝗮𝘁𝗶𝗼𝗻: 𝗙𝗿𝗶𝗲𝗻𝗱 𝗼𝗿 𝗙𝗼𝗲? 🌍 𝗚𝗹𝗼𝗯𝗮𝗹 𝗖𝗼𝗼𝗿𝗱𝗶𝗻𝗮𝘁𝗶𝗼𝗻 𝗡𝗲𝗲𝗱𝗲𝗱: Banks face fragmented regulations (UAE, UK, US), slowing adoption 🤝 𝗖𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗲 𝘄𝗶𝘁𝗵 𝗥𝗲𝗴𝘂𝗹𝗮𝘁𝗼𝗿𝘀: Sandboxes and tech sprints can bridge the innovation gap 𝘒𝘦𝘺 𝘵𝘢𝘬𝘦𝘢𝘸𝘢𝘺: 𝘭𝘢𝘳𝘨𝘦 𝘳𝘦𝘨𝘶𝘭𝘢𝘵𝘦𝘥 𝘧𝘪𝘯𝘢𝘯𝘤𝘪𝘢𝘭 𝘪𝘯𝘴𝘵𝘪𝘵𝘶𝘵𝘪𝘰𝘯𝘴 𝘢𝘳𝘦 𝘯𝘰𝘵 𝘢𝘣𝘭𝘦 𝘵𝘰 𝘣𝘦 𝘦𝘢𝘳𝘭𝘺 𝘢𝘥𝘰𝘱𝘵𝘦𝘳𝘴 𝘰𝘧 𝘯𝘦𝘸 𝘵𝘦𝘤𝘩𝘯𝘰𝘭𝘰𝘨𝘪𝘦𝘴 Thanks, Emma Johnson,Vilmos Lorincz, Ronit Ghose, Mikheil Nadareishvili and Amit Gupta for the great insights! #AI #FinancialServices #GenAI #Innovation #AIEthics #FutureOfBanking

"The o3-mini release proves that OpenAI is still in the fight - but this first post-DeepSeek round is, at best, a draw." Yiannis Antoniou, Lab49's head of AI, data and analytics recently shared his thoughts on OpenAI's two new variants, o3-mini and o3-mini-high, in an article for AI Business. With improved performance in coding, STEM, and math, the models offer free-tier access to advanced reasoning capabilities for the first time. However, DeepSeek still holds its own in terms of reasoning style. By revealing more of its thought process and maintaining a more human-like reasoning approach, it remains a compelling alternative. Read Yiannis's full article below. #AI #DeepSeek #OpenAI

Lab49's head of AI, data and analytics, Yiannis Antoniou, shared his thoughts in Verdict today on the release of the DeepSeek R1 family of AI models. R1 not only holds its own against the best offerings from OpenAI but also redefines the rules of the game with its open-source commitment and low-cost innovation. While debates about the actual cost are ongoing, the optics of achieving on-par capabilities with leading US labs at a fraction of the cost have caused a stir. This breakthrough has significant geopolitical implications as China has positioned itself to challenge U.S. dominance in generative AI. Read the full article below. #DeepSeek #AI #Innovation

Yiannis Antoniou, Head of Data, Analytics and AI at Lab49, recently shared insights in Unite.AI about how firms can bridge the gap between AI’s vast promise and its currently limited practical impact on enterprise operations. Organisations often struggle to move beyond experimental pilots to scalable, workflow-integrated tools. To unlock AI’s value, firms must focus on a product vision when designing AI tools, prioritise clean and well-governed data, and make Responsible AI a centrepiece of strategy. #AI #Innovation #ResponsibleAI

“OpenAI’s announcement of Operator, its latest foray into the agentic AI wars, is both fascinating and incomplete.” Yiannis Antoniou, Head of AI, Data, and Analytics at Lab49, recently featured in an article for AI News explaining how OpenAI's Operator streamlines AI interactions through web browsers, eliminating the need for complex APIs and bespoke integrations while enabling accessibility and ease of use. Read the full article in AI News below to learn more about OpenAI's latest announcement and Yiannis's analysis. #OpenAI #Operator