LLMs, RAG and AI Agents Author, "hands-on-RAG for production" (learning.oreilly.com/library…)
California, USA
- Tweets 2,846
- Following 1,317
- Followers 1,463
- Likes 2,994
Pinned Tweet
Yesterday I received my author copies of the "Hands on RAG for production" - it was very exciting to hold in my hand the actual physical copy, after so many months of work.
amazon.com/Hands-RAG-Product…
@OReillyMedia
5
372
Nightmare tweet for the modern software engineer:
“We are sorry that codex, Claude code, antigravity, copilot, cursor and opencode are down now.”
165
Ofer Mendelevitch retweeted
I am at @WeAreDevs congress in San Jose today! Find me if you want to make your Spark jobs 10x faster and 50x cheaper with @LakeSail
2
2
185
Agents are getting good at talking to each other. Sooner or later, one of them has to touch real code.
Today we're launching the BAND Python Kit for Docker Sandboxes, built with @Docker. One command runs a coding agent inside an isolated Docker Sandbox microVM and connects it to the rest of your agent team in a BAND room.
Picture a room with a hosted planning agent, a hosted research agent and a local coding agent. The planner decides how the change should be made, the researcher digs into the relevant API, and the coding agent makes the change, runs the tests against your workspace and @ mentions a reviewer with the result. Only the coding agent ever sees your repo.
Docker Sandboxes handles the boundary: its own kernel, filesystem and network, network access that's deny-by-default, and credentials managed through a host-side proxy instead of sitting inside the VM. BAND handles the collaboration: persistent agent identities, shared rooms, @ mention routing, message history, and every tool execution event visible to the team.
Docker gives your agents a safe place to run. BAND lets them work together from wherever they run.
2
2
2
104
Anyone can get an agent to build an app. Can you get a band of them to build it while you're not watching?
@WeAreDevs x BAND Hackathon, together with @Docker and @lablabai, taking place Sep 26 – Oct 5, online. Build an AI software factory in BAND Desktop. Two tracks, $6K in prizes.
Learn more here and sign up - it's going to be awesome!
lablab.ai/ai-hackathons/wear…
1
2
82
Congrats! Yes another great model from @AnthropicAI
Looking forward to using it!
Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family.
It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.
2
167
This hackathon, as part of the @WeAreDevs North-America event, is kicking off this Friday Sep 26; you to build a dark AI factory using @band_hq.
Five days until the lights go off.
Sign up for the @WeAreDevs x BAND hackathon and get a free ticket to WeAreDevelopers World Congress, San Jose, Sep 23-25.
The hackathon itself is fully online, Sep 26 - Oct 5. $6,000 across two tracks.
lablab.ai/ai-hackathons/wear….
2
186
We're sponsoring Day ZERØ's live Hack Day at The @AIconference this year. It's happening on Sept 29 and we cannot wait to be part of a full day of people building with AI agents - and seeing how they work together.
We'll be at booth 226 for the rest of the week if building isn't your day one plan.
2
3
101
Join the @WeAreDevs x @band_hq hackathon to experience first-hand the next agentic revolution: building an AI software factory.
The lights are off. The code ships anyway.
@band_hq leads the WeAreDevelopers x BAND Hackathon. Build an AI software factory: coding agents that plan, implement, and check their own work.
Sep 26 - Oct 5. Fully online. $6,000 across two tracks.
lablab.ai/ai-hackathons/wear….
1
117
Excited to be part of the panel tomorrow at the "Data for AI" event at @Snowflake 's SVAI in Menlo Park, along with others from @Intuit, @salesforce , @datastrato and WEX.
We’ll be discussing what needs to change in the data stack for AI agents to work with enterprise data in a governed, trusted, and semantically consistent way - a problem we think a lot about at @credibledatahq, through our work with open source Malloy.
RSVP here: lnkd.in/gy9ruZWn
1
69
I think we’re seeing a new discipline evolving, and I'm going to call it "Agentic Relations".
DevRel is about making developers want to understand, adopt, and build with your product.
But increasingly, your next “user” may be a coding agent. Can it discover your product? Understand what it does? Decide when to use you and for which use-cases?
1
73
Great post from Kyle.
The part that really resonates with me: we probably don’t know yet what the winning AI applications will look like.
And that’s exactly why builders need a trustworthy engine that understands the business, so they can build whatever comes next: agents, applications, dashboards, or things we haven’t thought of yet.
The first engines pumped water out of mines. As engines evolved, they unlocked railroads, cars, aviation, and spaceflight.
We think AI analytics is at a similar moment.
Credible is the AI Analytics Engine: you define what your data means, and Credible handles how it gets transformed, optimized, and served to agents, apps, dashboards, and people.
We build the engine. You build what comes next.
credibledata.com/blog/posts/…
120
Great article comparing how semantic modeling works: Malloy / @credibledatahq vs @getdbt.
noisetosignal.co/ideas/seman…
98
AI agents need governed enterprise data to work reliably, and I've been posting recently about how open-source Malloy is uniquely positioned to enable that.
So I took it to the test, with a complex dataset of security vulnerabilities, to explore how to connect @AnthropicAI Claude Code to Malloy and run the end-to-end flow: raw data → semantic model → governed definitions → analysis → working data app.
@ofermend took our open source skills and pointed the whole chain at 320,000 public CVE records + CISA's known-exploited list.
Raw feeds → governed model → shipped data app. No SQL reference book required to read the answer.
Microsoft: 853 critical CVEs, 368 known exploited.
credibledata.com/blog/posts/…
2
5
207
Thanks @sarthathak - it was a great and event. In order to run a zero-human company, one of the key capabilities is multiple agents communicating to each other, which is exactly what @band_hq provides.
Last weekend, we brought together 200+ founders, engineers, and researchers to compete for $12,000 in cash
The World's First Zero Human Company Hackathon by @tryterac challenged buildings to create an autonomous company, capable of running on its own
Had an amazing panel with @DStrachman, @aridutilh, @shilpinayak, led by @gabriella_12678
The judging panel was insanely stacked, with 5 @ycombinator founders, multiple @zfellows, and staff from @stripe & @GoogleDeepMind.
Big thanks to @stripe, @Lovable, @whop, @thelinqapp, @replayio, @superserve_ai, @fastinoAI, @band_hq, @getperflo, and others for sponsoring!
This was just the start of the biggest series of zero human company events in the world.
For the next year, we'll be hosting these across globe.
Which city do you want us to come to next?👇
2
223
Excited for this hackathon this weekend to see how folks will use @band_hq to build the world's first zero human company.
This Saturday, we're giving out over $10,000 at the World's First Zero Human Company Hackathon with @stripe, @Lovable, @whop, and others
You'll have 12 hours to build an agent that earns real revenue and pitch to founders of @ycombinator and @zfellows companies, and judges from @GoogleDeepMind and @OpenAI.
We're bringing top investors from the @1517fund, @solofounders, other VCs
This room will be filled with the next generation of AI-native founders, engineers, and researchers.
We're still accepting registrations for the next 36 hours, so comment if you're interested and I'll send you the RSVP
Thank you to all of our amazing sponsors! Render, Linq, @superserve_ai, @replayio, @fastinoAI, @band_hq, @dodopayments, and others
3
139
Ofer Mendelevitch retweeted
Big news: @credibledatahq is now on the Claude community marketplace.
@ClaudeDevs can now bring governed semantic models - and trusted business context - directly into their Claude Code workflows.
Two simple steps inside Claude Code:
1. "/plugin marketplace add anthropics/claude-plugins-community"
2. "/plugin install credible@claude-community"
1
1
58
We had a great time at @Ai4Conferences last week! As well as hearing some of the amazing speakers, including our very own CEO @g00manoid, we got to meet so many of you at our booth - thank you for coming to visit and hear more about how we're bringing agent collaboration to development teams across the world. Plus, we beat that pesky ORChestrator back... #collaborationFTW
1
5
127
Check out my latest article: Your LLM Isn’t the Bottleneck. Your Retrieval Pipeline Is. linkedin.com/pulse/your-llm-… via @LinkedIn
2
95
Today's highly anticipated keynote at @Ai4Conferences did not disappoint, with @drfeifei @AndrewYNg and @geoffreyhinton discussing topics they don't really agree on: commercial vs open models, AI regulation, and the impact of AI on jobs.
Two of my favorite parts:
1. There was a lot of discussing about "fear mongering" in AI (which of course Geoff did not agree with), but I liked FeiFei's statement about how to drive rationale discussions about this : "let's bring science, not science fiction"
2. Geoff made this point about copyright: the big commercial labs say that (I'm rephrasing a bit) "we can't negotiate a deal with every content creator out there - it's just non-scalable; but I (Geoff) think that it's actually quite achievable with agent to agent negotiations". I totally agree and in fact that's exactly what we're building at @band_hq.
1
3
180