We released MCP server functionality! You can now use the your data safely and securely with Cursor, Claude, and OpenAI
Try it out here https://t.co/Xd2oTXr9PZ
Excited to welcome @vcastellm as our technical co-founder and CTO. Victor brings 20 years scaling distributed systems under real pressure. He is the creator of Dkron, now running in production at American Express and Flickr. The technical depth this problem requires!
We're building compliance-native RAG infrastructure for regulated industries — where data isolation, audit trails, and role-based access controls are the foundation, not an afterthought. Follow Victor's journey building in the open.
Need a demo? https://t.co/nzHjgqbIhg
Hey everyone, I have some news to share:
Four years ago I joined Polygon, coming off a long tenure at Jobandtalent where I worked with one of the best engineering teams in Spain. I grew a lot there scaling systems and databases to millions of users and bringing Web 2 best practices into a fast-moving company. Then I wanted to go deeper into distributed infrastructure and brought that experience into the Web3 ecosystem at Polygon. Over time I got into protocol internals and client development. The most rewarding chapter of my career so far.
As you may know, everything comes to an end. Back in mid-January, there were layoffs at Polygon https://t.co/wy96kpnEIX.
I took some time to reflect on what I actually wanted to do next. My timeline has been on fire since the releases of Claude Opus 4.5 and Codex 5.3, no surprise, since I’d been delivering features at Polygon without writing any code since November. The intersection of AI and infrastructure had been pulling at me.
Then came the right timing. Matthew Rossi, a friend and former colleague at Polygon, reached out. His team had built something interesting, and uncovered an even more interesting problem underneath it.
This is why I am genuinely excited to announce that I am joining Matthew and Katara AI as technical co-founder.
At Katara AI https://t.co/oNWwAI7UqL, we are building compliance-native RAG infrastructure for enterprises in regulated industries. The specific problem: most AI infrastructure was built for speed, not for the governance questions that arrive after deployment, what data entered the system, where it went, who had access, and whether you can prove it to an examiner. AI adoption in financial services, legal, and healthcare is moving faster than the infrastructure those industries actually need. That gap is worth fixing.
I am not leaving the Web3 world entirely, I will continue advising projects I believe in. But this problem is too important and too well-timed to ignore.
We will be building in the open. Follow along as we navigate what it actually takes to build compliance infrastructure for the AI era.
If you are working on enterprise AI deployment, regulated infrastructure, or just want to follow the journey, my DMs are open.
Data Quality is THE #1 Hurdle for AI in 2026!
Companies are battling poor data, leading to unreliable agents & hallucinations that kill trust, amp errors, and stall scaling beyond pilots. Without grounded, business-specific data, AI can’t deliver facts. https://t.co/Lq5ZO66Oow
Today, we’re launching Katara’s Model Context Protocol (MCP) Tools!
Load your data into Claude, Cursor, & ChatGPT faster than ever.
Prep data for agentic apps in a snap!
Try it: https://t.co/76uw93Tkjx
We released MCP server functionality! You can now use the your data safely and securely with Cursor, Claude, and OpenAI
Try it out here https://t.co/Xd2oTXr9PZ
Katara is setting the standard for deploying Agentic RAG
The system automatically indexes files, documents, and data to fuel diverse use cases with LLMs and AI Agents, ensuring high-quality outputs through RAG fast retrieval, automatically indexes, and production-grade optimizations.
Get started: https://t.co/r7brbnTL1p
Imagine your BEST Community Manager or Dev Rel
1️⃣Able to answer every question
2️⃣Across every channel
3️⃣Perfectly in under 7 seconds
How much development and growth would that unlock for you?