Deep in Agents, LLMs and automation · Open source at ♥ · Sharing AI gems daily to help you work smarter · Ex-data scientist @Snowflake | @Streamlit ·
📩 DM for Collabs
Joined January 2009
- Tweets36K
- Following49.5K
- Followers183K
- Likes11K
Charly Wargnier ♨️ retweeted
these guys just dropped the complete AI product launch playbook for 2027 🔥
save this.
Charly Wargnier ♨️ retweeted
If you want to really understand AI agents, spend some time on this new site.
You won't regret it.
secondbrainos.dev turns the whole subject into a map you can explore, and it's honestly brilliant.
Every page, term and tool is a bubble.
The lines show what depends on what, so you can actually see how the pieces fit together.
here's what's inside:
> a 7-module course, from a single prompt to a production agent
> 5 handbooks, each exploring one layer and ending with a build you can finish in an evening
> a 65-page hands-on project where an agent builds and maintains a second brain for you
> a glossary of 57 terms and 100+ tools
130 pages. 1,236 connections. 9-10 links per page.
It's FREE, written in plain Markdown, and the repo (in 🧵↓) is public too.
Charly Wargnier ♨️ retweeted
BAIDU JUST DROPPED A MAJOR RELEASE FOR OPEN-WEIGHT AI
and this one gives devs way more control over what comes next.
🚨 Meet ERNIE-5.5-Flash-Base.
@Baidu_Inc has released the pretrained foundation model BEFORE any mid-training or post-training.
Why is that a big deal?
Training a foundation model from scratch takes HUGE resources.
Building on someone else’s finished model means inheriting their later training choices.
this release gives devs an earlier starting point, with uber strong capabilities already in place.
Baidu reports SOTA performance overall among the similarly sized base models it evaluated.
Some concrete results against DeepSeek-V4-Flash-Base:
> HumanEval+: 82.3 vs. 69.5
> MMLU-Pro: 78.4 vs. 69.2
> LiveCodeBench: 40.1 vs. 31.9
All before mid-training or post-training
Let's dive in 🧵↓
Charly Wargnier ♨️ retweeted
This is the real bottleneck.
Coding agents ship apps fast. Reliable data + citations on messy real-world data is harder.
Been fighting the same thing in financial RAG (SMRA) — Indian filings still hallucinate numbers.
Grounding > model size.
Coding agents can ship an app in hours.
A reliable data pipeline? that's another story!
Data is messy. constantly changing. and unforgiving.
@Eon_io_ just launched Yukon to change that.
> ingest, process, and serve in one flow
> open Iceberg tables in your own bucket
> 4x less compute than Spark, 7x less than Flink
Charly Wargnier ♨️ retweeted
LOCAL AI JUST GOT A LOT MORE INTERESTING 💪
Choose your models.
Run them on your own hardware.
Put them to work with your agents.
@atomic_chat_hq's new Local AI Server brings it all together:
> install an engine matched to your hardware
> download compatible models from Hugging Face
> connect agents like Hermes, OpenClaw or Codex
> manage everything through a CLI or web UI
it supports llama.cpp, MLX, vLLM and more, with an OpenAI-compatible API to connect your tools.
Less setup to piece together.
More time to explore models and build with them.
Free. Open source. Running on your own hardware 🔥
Introducing Atomic Local AI Server ⚡
Choose inference engine. Run local models. Connect your agents.
• Multiple engines: llama.cpp, vLLM, SGLang, TensorRT, MLX, Prism and more.
• Local AI server with an OpenAI-compatible API.
• One-command engine installation matched to your GPU, CPU or NPU.
• Pull from 3M Hugging Face models in GGUF, MLX or safetensors formats.
• Run on Windows, Linux, Apple Silicon and NVIDIA hardware, including RTX, Blackwell and DGX Spark. AMD Ryzen Halo support coming soon.
• Connect Hermes, OpenClaw, Claude, Codex, Pi agent and others.
• Generate images and videos with stable-diffusion.
• Manage everything through CLI or web UI.
• Remote access via Cloudflare tunnels
Free. Open source. Local. 👇
github.com/AtomicBot-ai/atom…
Charly Wargnier ♨️ retweeted
WHAT?! SOMEONE JUST OPEN-SOURCED TERRAFORM FOR AI KNOWLEDGE GRAPHS 🤯
It's called Open Ontologies and it makes AI reasoning verifiable.
Here's the crazy part:
→ Change 1 line, uncover 901 hidden consequences in 0.04s
→ Verify 259 facts, reject a forged one
→ Preview the blast radius before applying changes
→ Generate proofs anyone can independently check
→ Validate RAG claims against their sources
Works with Claude Code, Cursor, Windsurf, VS Code and more via MCP.
One Rust binary. Runs locally. Fully open source.
100% Free and open-source.
Repo link in 🧵↓
Charly Wargnier ♨️ retweeted
🚨 NVIDIA JUST OPEN-SOURCED A SEMANTIC LAYER FOR AI AGENTS
It’s called Auto Ontology, and it tackles a problem that better SQL alone won’t fix.
An agent can query the right database and still return the wrong answer.
Why?
Because “revenue” might mean something different in every system.
Auto Ontology is a Python framework that builds shared context over enterprise metadata:
→ Entities and business metrics
→ Relationships between data
→ Certified definitions and lineage
→ Provenance behind answers
It can generate ontologies from metadata, connect existing semantic models, and reconcile conflicting definitions.
For structured data, it also supports natural-language-to-SQL with validation against the ontology.
Agents can access that context through MCP, CLI, or API.
The useful shift: giving agents access to business meaning alongside the data.
Open source under Apache 2.0.
Repo link in 🧵 ↓
Charly Wargnier ♨️ retweeted
50X FASTER AND 100X CHEAPER THAN LEADING VIDEO MODELS 🤯
Atmee v2 is a video foundation model built for developers.
Give your coding agent an API link, describe a character, add an image and voice clip.
Test it. Ship it. Done.
🤝 Paid partnership
After Google acquired my AR startup, I started thinking about the next interface, this time for AI. Text and voice work fine, just like DOS did before the Mac. But nothing’s more intuitive than face to face.
After two years of research, today we’re launching Atmee v2: a real-time avatar model that makes live video agents feel truly life-like.
• Any character from 1 image + 10s of audio
• Contextual expressions, gestures, even interacting with the environment
• Live video calls, voice only, or video generation
At its core, Atmee v2 is a video foundation model, it can render anything, but it’s built to excel at characters, which is what video models like Seedance 2.5 and Wan 3.0 are mostly used for. In comparison:
Atmee v2
• Up to 50x faster
• 10-100x cheaper
Build tutors, sales coaches, support agents, companions, game characters, even whole living worlds. The interface after text and voice starts today.
Charly Wargnier ♨️ retweeted
WAIT.. SOL DOES ALL OF THAT FROM AN EMAIL?
Sol handles all the prep:
> finds the task
> gathers the context
> gets the work ready for review
You just review and approve.
stupidly useful.
🤝 Paid partnership
We shipped a BIG update to Sol (hellosol.app)!
Sol is for people who have consequential work in their inbox.
The most fun part tho - was watching the bets behind this update play out with early users:
- 100+ new emails a day - is the norm for a LOT of people: founders, execs, sales, recruiters, consultants, ops.
- To make things interesting - every email asks for something different: a decision, an approval, a meeting, a proposal.
- The heavy lifting is in: arriving at the decision, doing the review, picking if/when to meet, creating the proposal.
- Drafting the reply - is merely the outcome of that heavy lifting. That alone doesn’t solve the real problem.
That’s where Sol is different. It works on every email as soon as it arrives, does the heavy lifting each one needs, and prepares it for your approval.
You approve or edit. It learns each time and keeps getting better.
When your inbox is filled with highly consequential work, you need an assistant that’s doing the heavy lifting, continuously.
That’s what we’re building Sol to be - proactive about what you need, 100 times a day, every single day.
Extending early access to this big update ✌️
Charly Wargnier ♨️ retweeted
STOP RENTING A GIANT MODEL FOR EVERY SMALL TASK
Ticket routing. Intent classification. Repetitive decisions.
Oumi's FAiE helps turn those jobs into specialized open models, complete with evaluation, fine-tuning and deployment.
Cos' not every workflow needs frontier-scale AI ;)
🤝 Paid partnership
We're using AI agents to automate everything. Except the engineering of AI agents themselves.
That seemed backwards to us.
What if you could train your own specialized AI agent with just 15 minutes of your time instead of months of AI engineering?
Would you keep renting the same expensive generic models as your competitors, endlessly patching prompts and harnesses until you hit a wall?
Today, we're launching the agent that builds, trains, and continuously improves AI agents. 🚀
Meet FAiE, our Frontier AI Engineer. Describe what your agent should accomplish, and FAiE designs evaluations, diagnoses failures, curates data, selects training strategies, and runs experiments.
15 minutes of your effort. A few hours of automated AI engineering. Your own specialized agent, with its model weights trained for your task.
Already have an agent? Keep your harness. Connect it to @oumi_ai and train the model behind it with SFT, DPO, or GRPO (Private Preview).
The agent that builds your best agent.
Excited to see what you build: platform.oumi.ai/
Live demo today at 10 AM PT: oumi.ai/events/fall26
Charly Wargnier ♨️ retweeted
HOLY MOLY. BUILDERS ARE GOING TO LOVE THIS
X’s Head of Design @benjitaylor just shared a skill that teaches AI agents his product demo process.
Stageshow gives your agent instructions to create:
→ Product UIs that look like real apps
→ Dashboards, feature cards, and phone flows
→ Bento grids for landing pages
→ Animated clicks and smooth transitions
→ Full product walkthroughs
The wild part: it watches its own recording frame by frame and fixes the rough edges.
The output is plain React. Yours to edit.
Free and open-source.
Repo link in 🧵↓
Charly Wargnier ♨️ retweeted
No f*cking way
now you can just tell DeepSeek Harness what’s you want…
a sandbox game, a 3D LLM visualizer
..and it builds the plugin 🤯
Charly Wargnier ♨️ retweeted
Today, we're launching @genproof: a research lab making formal methods accessible.
We found bugs in 100+ OSS projects, including 6 libraries “formally verified” by Anthropic. We’re also top-1 on the lean-eval leaderboard.
With our tools, you too can become a verifier: 🧵
Charly Wargnier ♨️ retweeted
WAIT... THIS USED TO TAKE 10+ MODELS 🤯
@hume_ai says its new Audio Decision API brings that down to one call.
Count speakers. Detect laughter. Ask whether a caller sounds frustrated etc. etc.
Same recording. Multiple questions. One API call 👊
🤝 Paid partnership
Introducing the Hume Audio Decision API.
Ask anything: How many speakers? Is anyone laughing? Is the caller frustrated?
What used to take 10+ models is now one call.
Trained on 10M+ human judgments, so it answers the way human listeners would.
Realtime, or across hours of recordings.
Request access below.
Charly Wargnier ♨️ retweeted
THIS IS BASICALLY A FULL 3D PRODUCTION STUDIO INSIDE YOUR CODING AGENT.
And it's open-source.
@Scenario_gg just released 3D OS:
→ 6 specialized departments
→ 80 production-ready skills
→ Direct integration with professional 3D tools
Here's the breakdown:
📁 Animation → Rigging, motion capture, video-to-motion
📁 Art → Characters, creatures, previs, storyboards
📁 Modeling → Image-to-3D, hard-surface modeling
📁 Look Dev → PBR materials, texturing, lighting
📁 Worlds → Walkable environments, skyboxes, splat scenes
📁 Pipeline → Retopology, UV baking, game-ready export
Each department handles a different stage of production, passing outputs to the next.
Even better:
Your agent can operate Blender, Maya, ZBrush, Unreal, Unity, Godot, and Roblox locally.
One prompt → 3D production pipeline → game-ready assets.
Runs inside Claude Code, Codex, or Cursor.
Free and open-source.
Repo link in 🧵↓
Charly Wargnier ♨️ retweeted
THIS IS HOW OBSERVABILITY SHOULD WORK
Your logs stay in your cloud.
AI finds the issues.
You own your data.
Sazabi is absolutely crushing it again 🔥
Today we're launching Sazabi BYOC 🔒
The part of @sazabi that receives, processes + stores your logs now runs in your cloud
