Ran Tavory

Kiro helps teams move from intent to production-ready code using structured, spec-driven workflows. From today, it can also reason with live runtime context by integrating with Lightrun MCP. That means Kiro can now see how its code actually behaves at runtime and use that context to design more reliable changes, investigate issues, and validate fixes. 📽️ Watch Barak Ben-Ari, Product Manager at Lightrun, demo how Kiro uses runtime context to investigate issues in a running trading application, live This is reliability by design.

38

‏6‏ ‏תגובות‏

🚀 Excited to be heading to AWS Summit Tel Aviv this month! Always a great opportunity to catch up with people from the tech ecosystem, exchange ideas, and discuss what’s actually happening with AI beyond the hype. At Reverso & Ginger Software, we’re working on a wide range of AI and language technologies - from machine translation and AI-powered writing assistance to voice AI, intelligent document processing, LLMs and AI at scale. I’m also particularly interested in exchanging around MLOps, FinOps, and AI cost optimization, and more generally, how to turn AI capabilities into reliable, scalable products discoverable and used by millions of people. If you’re attending and working on any of these topics (or just feel like comparing notes over coffee) drop me a DM and let’s meet! ☕️ 📍 Expo Tel Aviv 📅 September 10, 2026 #AWSSummit #AWSSummitTelAviv #AWS #AI #GenAI #MachineTranslation #MLOps #FinOps

14

‏3‏ ‏תגובות‏

It's not easy to learn a new language. One of the best ways is to just read. But what happens when you don't really understand everything that you read? You translate it. So I built a small app that automates this for me - it displays the original (German, in my case) text side by side with the English translation, sentence by sentence. Just paste a URL, and get a clean bilingual reading view. For me it's super helpful to improve my German skills. And yes, there are apps that do this, including Chrome extensions and paid mobile apps, but it was easier to build my own custom solution (link in the comments). Built with Next.js, Google Translate, and Claude Code - which wrote most of the code. Deployed on Vercel.

8

‏1‏ ‏תגובה‏

Did you know that even the most advanced language models make mistakes in nearly 37% of cases? Prof. Amnon Shashua mentioned this just now in an interview at Reichman University conference on AI Opportunities, Challenges, and Policy. Most of you probably aren’t surprised. But the real question is: does this statistic make us double-check the information every time we use these tools? Not necessarily. For organizations, this highlights a critical point: the need to rely on systems that analyze only the organization’s internal knowledge, surface the most up-to-date documents, give always the exact reference and build intelligent layers on top of language models to minimize these gaps. That is crucial part of what we do at Xarcam Just saying. ‏‏Elad Kleinbord‏ ‏Ben Kleinbord‏ #AI #AIinorganizations #innovation

10

‏1‏ ‏תגובה‏

When as the last time you merged a PR to Google ?! Because our very own Ari Heber just did! 👏 As we see growing usage of Yuki's BigQuery, our customers wanted to increase their usage in various AI workloads, including GoogleAPIs MCP-Toolbox, so we didn't just wait, and opened an ambiguous sporadic Feature Request, we opened a Pull Request straight to their code. Great to see companies and projects like these open their source-code to external contributions, accelerating the growth together.

39

‏2‏ ‏תגובות‏

Architecting the Agentic Organization: MyHeritage’s AI Journey Scaling AI across an enterprise engineering team requires more than just writing faster code — it demands a complete architectural and cultural shift. Discover how MyHeritage transformed its engineering lifecycle, automated quality gates, and integrated autonomous agents to achieve a real 3X productivity multiplier. Here is how MyHeritage’s AI-driven transformation measures up against the core principles of "The Agentic Awakening · An AI-Native Engineering Playbook." Link to the full article in the first comment.

60

‏10‏ ‏תגובות‏

AI is no longer a side topic for technology leaders. It is changing how organizations design products, run operations, build software, manage risk, and think about governance. At DevGeekWeek 2026, I’ll be leading a seminar for managers and technology leaders on building AI-native organizations: strategy, architecture, agents, MCP, automation, development, QA, security, FinOps, and the future role of engineering teams. The seminar offers a practical, structured view of the strategic, architectural, operational, and professional questions technology leaders need to address now. You are welcome to join. Seminar details: https://jbh.co.il/l/1yy

20

AI makes development faster, but also messier. Behind every accelerated task, it’s planting quiet time bombs 🤯: duplicated code, missing safeguards, subtle vulnerabilities that erupt later in production. My new article on The New Stack details the four most common time bombs, and how to neutralize them. https://lnkd.in/dm4m3j-7

18

‏4‏ ‏תגובות‏

Product update: Islo, the AI sandbox by Incredibuild, is LIVE! 🚀 “AI coding agents promise 24/7 dev cycles, but they’re falling short exactly where R&D has failed for years: CI/CD pipelines. These days are over.” - Shimon Hason, Incredibuild CEO Read the full story: https://lnkd.in/djk9bEwT #AICodingAgents #AISandbox

35