Exadel

Today we had a Town Hall for our Performance Managers (Exadel), where we introduced our new model for working with and supporting employees. While preparing for the meeting, I caught myself thinking about something I had probably never articulated this clearly before: We’re very good at knowing what we expect from the company. But we think far less often about what the company expects from us. And I mean something much broader than doing a great job on a client project. As usual, the easiest way for me to explain this is through CrossFit. I’ve been training with the same group for around 11 years. I train with them, write workouts, and coach them. And there is one rule I never allow people to break: You cannot keep taking from the group without giving anything back. That’s why I don’t like it when someone comes to a group workout, puts on headphones, and completely disconnects from everyone around them. Technically, they’re doing the same workout. But a group is not just a program written on the board. It’s the people next to you. The shared energy. Support. Competition. Jokes. Collective suffering. And genuine support for the person who finishes last. If you benefit from all of that while staying on the outside, the relationship becomes one-way. The same applies when someone comes to a group workout, spends half of it following their own program, and joins the group only for the parts they happen to like. CrossFit is incredibly scalable. You can substitute movements, reduce weights, and adjust intensity. But there’s a fundamental difference between adapting a workout to your capabilities and using the group as convenient infrastructure for your own program. And the more I think about it, the more I realize that companies work in a very similar way. We join a company with perfectly reasonable expectations: a good project, professional growth, stability, clear employment conditions, and support when something goes wrong. There’s nothing wrong with that. But there’s another side to the relationship. The company has expectations of us too. It’s easy to understand that a client expects something from us. But sometimes we forget that the client is the client - and the company is the company we’re part of. Clients can change. Projects end. Assignments change. But your relationship with the company continues. So contribution can’t end with: “I did a great job on my project.” There is another question worth asking: What became stronger inside the company because I work here? And there’s one more uncomfortable question behind it - in the first comment.

The hardest part of leadership isn't making difficult decisions. It's making decisions when there isn't enough information, there isn't enough time, and certainty simply doesn't exist. Eventually, every leader faces that moment. The discussion stops. And you realize that waiting for certainty is no longer an option. Leadership isn't tested when the path is obvious. It's tested when every option carries risk. You choose a direction. You explain why. You commit to it. And if new facts emerge, you have the courage to change course. Because in my experience, the cost of indecision is often higher than the cost of making an imperfect decision. I recently had the opportunity to reflect on this and other leadership lessons during an interview for Exadel (Women@Exadel), where we discussed growth, resilience, and the realities of leading through uncertainty. If you'd like to read the full interview, you can find it here https://lnkd.in/dvDUsdZ4

Today, I'm proud to welcome Tangent to Exadel . This isn't just an acquisition. It's a strategic move that closes the loop on how enterprises experience AI: from vision and design all the way through engineering and delivery. Tangent brings world-class AI-driven digital strategy, consultancy, and product innovation. Combined with Exadel's global engineering depth and offerings like Exadel Colleague, we can now deliver end-to-end AI transformation that creates real competitive advantage, not just incremental improvement. We're also expanding our footprint in the UK and Europe, and bringing extraordinary talent directly into client engagements. The pace of change is accelerating. So are we. Welcome to the family, Tangent. Let's build something remarkable. #Exadel #Tangent #CompanyCulture #Acquisition #Growth

You can’t have enterprise AI without enterprise data governance. It sounds obvious, but it’s the number one thing slowing down AI adoption right now. I’m heading to TechEx Events AI & Big Data Expo in San Jose with the Exadel team, and I’m looking forward to talking about building an AI-ready foundation, not just a collection of cool tools. Curious about your own data maturity? You can benchmark it right here with our AI Maturity & Readiness Assessment: https://lnkd.in/eDFBpm5d Don’t miss the Day 1 Panel: “From Assistants to Autonomy - Architectures Enabling Next-Gen Agents,” where Exadel’s Devendra Sharma will be covering decisions that separate agents that wow in a sandbox from agents that hold up under real workloads. If you're attending, let’s talk about your data infrastructure and how to clear the path for production-ready AI. 📅 Book a time here: https://lnkd.in/eHRuW3mi If any of my connections are headed to the expo, it would be great to catch up. Stop by Booth 229 and say hello! #DataGovernance #BigData #AIReady #TechEx #Exadel

BTW, here is my Exadel Code Fest 2026 project. Yep, just another AI-chat-bot from the first sight 😅 . But the difference is in the details! 🤓 * I'm incorporating Karpathy's trending "LLM-Wiki" data model, which gives a great precision of data retrieval without a complex RAG pipelines. 📖 * OpenCode based agentic harness with self made orchestration layer 🤖 * MCP and skills for additional data sources integration (Gitlab/Github) 🔄 * Architecture made for a large organization level scale out of the box. Microservices, background async jobs, data caching, cloud-native infrastructure compatibility. 🦾 All this features make a "just-another-chatbot" to be a production level product, which has already started integrating into #Exadel real projects! And it already helps our team to get up-to-date project data, people and decisions information, be in the stream of product development with regular tech digests, announce PR's details and notify DevOps on CI/CD failures. Taking 1st place on Hackathon is not so important for me, but integrating software into real projects and see how it really helps us in software development - this is what makes me thrilling! #Exadel #CodeFest2026 #AIDevelopment #Agents #Harness

Does your engineering team feel like AI is a bottleneck instead of an accelerator? I’ll be on the ground at TechEx Events in San Jose, CA with the Exadel team of Devendra Sharma , Sugra Naqvi , John Madden to help you pinpoint what's breaking in your AI roadmap. We’ve put together an AI Maturity & Readiness Assessment that touches on exactly these technical bottlenecks. Check some details here: https://lnkd.in/dwwZzgdG Meet me at Booth #229 and let's chat about scaling your engineering operations. 📅 Book a slot with our team: https://lnkd.in/dGpxGegd #EngineeringLeadership #AIatScale #TechEx #Exadel #GenAI

AI adoption is hitting a physical limit. For the past year, the conversation has been dominated by what AI can do. Now, the conversation is shifting to what AI requires to function at scale. According to recent Goldman Sachs research, we are entering a selective phase where the "winners" aren't just those with the best ideas, but those with the best foundations. Why does this matter for enterprise leaders? 1. Supply Chain Complexity: Building the data centres needed for AI takes years, not months. 2. The Energy Equation: Power demand is skyrocketing. If you don't have a stable energy and cooling strategy, your AI growth will stall. 3. The Infrastructure Gap: We are seeing a 175% projected increase in power demand by 2030. As I often say, you can't scale AI on top of complexity and inconsistency. Whether it's your data foundation or the physical hardware it runs on, the "basics" are becoming the most competitive advantage a company can have. If you are looking to move from pilot to production, it’s time to look beyond the algorithm and start looking at the grid. #DataLeadership #AIInfrastructure #DigitalTransformation #Exadel #CDO

AI knows the spec. If your value is just that, you're already replaceable. Hot take: AI can read every FIX ( Fixnox ) tag, parse every Swift message, and calculate yield to maturity faster than you. And it still doesn't operate your system better than you do. Here's why: AI reads Tag 150 = ExecType "Filled" and moves on. → You know this client requires an ESG Book compliance review before confirming settlement. AI generates a valid pacs.008 ISO 20022 message. → You know a beneficiary in that jurisdiction triggers an internal KYC rule, not documented, learned from a past regulatory incident. AI fetches a Bloomberg price feed and maps it correctly. → You know the issuer was downgraded yesterday, and the pricing model needs recalibration before market open. The spec is public. The exceptions are earned. But here’s the shift most people miss: → Those exceptions won’t stay tribal knowledge forever. → They will be captured. Logged. Turned into rules. Embedded into systems. And eventually, learned by AI. So the real advantage isn’t just knowing the exceptions. It’s being the engineer who: → identifies them early → understands their impact → and turns them into systems before anyone else does AI is the tool. Domain knowledge is the starting point. Systems thinking is the real moat. AI (spec) + You (exceptions) + Systems (scale) = 10x fintech engineer. What exception in your system only exists because no one took ownership to turn it into a rule? #java #fintech #fixprotocol #swift #iso20022 #ai # Bloomberg #trading #backend #domainknowledge #Exadel # Fixnox