Linux Dev | AI Rev Early Adopter | Gen AI\UX | Deep Agents
Joined July 2023
- Tweets 3,962
- Following 543
- Followers 1,860
- Likes 37,747
631 Photos and videos
Side by side example
Same model (claude-opus-4-6). Same task. Two different agent harnesses
@LangChain Deep Agents CLI: 9s
Claude Code: 16s
The harness IS the performance. 1.7× difference, zero model changes
18
33
373
113,683
Building agent-native is here
In a short time there will be more agents visiting your app than humans, and they dont agree on the display layer
It’s interesting to watch this evolve real-time with new agentic-first systems coming online
Jun 4
the default experience for new software is now:
- point agent to good onboarding docs
- everything is headless and discoverable by the agent natively
- agent drives the interaction, human reviews and interjects as needed
but there's a large chunk of existing software that's slow moving to a Headless first world but is still important so we have to have some way to access it
Browser Agents are filling the gap, and they're getting better and better
this is because everyone's desired UX for software is Agent First. and ppl are building the tooling and infra to make what's actually happening feel like that desired UX
basically the world is building infra to patch any gap in an Agent First software experience and a lot of that infra is basically shaped by what agents are good at natively
the inherent strengths and weaknesses of agents are literally shaping what ppl are building
1
2
206
Neat - @GoogleAI just mentioned @ContextRepo along side one of my hero’s @AndrewYNg
A couple weeks ago I attended Andrew’s and @hwchase17’s talk about the Future of AI Agents live at LangChain Interrupt -
Inspiring 🍀
3
157
Git Maxd retweeted
May 26
fde's are genuinely just normal software engineers but with a little bit of rizz
23
6
259
31,376
Ridiculously cool open source example using @LangChain_JS
The Multi-Modal children’s bedtime story generator is an excellent example of LangChain streaming UI & stream namespace subscriptions - great patterns to learn from!
Great share @bromann 🔥
May 26
As an upcoming parent, I keep imagining all the tiny routines ahead 😊👼
One I am especially excited for: using AI for making up bedtime stories.
So I built an app that creates one on demand: story, illustrations, and narration all streaming in together with @LangChain_JS ❤️
3
2
7
2,880
Git Maxd retweeted
May 20
a few months back, it become clear to us that a large part of technical work would be driven by agents in the future. coding agents were becoming ubiquitous and highly capable.
since we build a platform for technical users, we needed to update our beliefs and strategy accordingly! LangSmith Engine automates the improvement of agents by looking through recent traces and finding problems according to a taxonomy of common agent issues that we have defined.
we launched the product at our annual conference last week and the reception so far has been very exciting. and we're just getting started 📈
3
2
17
3,415
AX - Agent Experience, is a thing you can start to optimize for today
Ora.run gives you the playbook, and the prompts to improve your AX
Developed by @assaf_elovic, the creator of @tavilyai
You’ll be surprised at all the low hanging fruit
2
4
413
Wow! What a week!
Lots of new things to learn - new ideas to form
The complete Agent Development Life Cycle 🎯
❤️ SF but can’t wait to get home and back to building!
Thanks for the warm welcome and fun After Parties!- Till next time! @hwchase17 @amadaecheverria @torres_andres87 @PetralliLucas @lgesuelli_p 🫡
May 15
ICYMI:
1️⃣ LangSmith Engine
2️⃣ SmithDB
3️⃣ Managed Deep Agents
4️⃣ LangSmith Sandboxes: Now Generally Available
5️⃣ Context Hub
6️⃣ LangSmith LLM Gateway
7️⃣ Sandboxes, Prebuilt agents, free model usage in LangSmith Fleet
8️⃣ Deep Agents 0.6
9️⃣ LangChain Labs
langchain.com/blog/interrupt…
3
9
20
7,673
Git Maxd retweeted
May 14
LangChain Labs is live
3
21
2,773
Bro is based 🔥
May 15
met the man!! big year let’s cook 🚀
3
232
Got to meet the Legend today!
Had a great talk with @Vtrivedy10 about Deep Agents, life and what is coming our way -
This will be a BIG year for @LangChain 🔥
1
2
19
4,315
Great opportunity to try out @LangChain Deep Agents and Fleet
Free tokens from @FireworksAI_HQ 🔥
- Save your API spend, learn for free!
May 14
We're offering free tokens in Fleet powered by @FireworksAI_HQ for all Developer & Plus plans
It's now easier than ever to get started in Fleet. Oh and did I mention you can now give all your agents access to a sandbox? It's like Christmas!!
2
5
417
Git Maxd retweeted
May 12
Interrupt! Tomorrow!
1
1
3
233
Rly looking forward to this Deep Agents workshop at @LangChain Interrupt
Learning everything I can about Context Management and Deep Agents are the best examples of OSS best practices
Harness to Fleet, Deep Agents are the way 🔥
Be in SF Tomorrow! Cant wait to see my people!
3
2
13
1,869
It’s true
Every indie without a token budget is figuring this out at the same time
And it’s glorious - OSS FTW
May 7
your daily reminder that open models are plenty capable for a lot of coding work. easiest place to feel that out is deepagents! swap the model and go.
i've been enjoying GLM-5.1, Kimi K2.6, MiniMax M2.7, DeepSeek V4 Pro.
here's some examples using our CLI agent in headless mode
1
2
170
Great new drop by @LangChain’s @hwchase17
“If you do not know what the agent saw, what it did, and what happened next, you cannot reliably know what to improve”
Traces give agents this insight and that’s a powerful thing 🔥
1
2
5
1,415
More model gold from @FactoryAI with this eye-popping drop
Starting to see a pattern - "Open-source models hit 85% of frontier accuracy at 1/3 the cost. At that price point, you can run multi-pass review and still come out ahead."
Similar results with @langchain Deep Agents
x.com/Vtrivedy10/status/2049…
Apr 29
Which model reviews code best?
We benchmarked 13 models on AI code review across real PRs and the results are surprising.
Spending more tokens did not result in better code review.
A $1.25/PR model beat another that was more than 2x the cost. Meanwhile, budget models at $0.15/PR delivered ~80% of the quality of frontier models while being 10-30x cheaper.
In fact, cost only explained ~21% of the difference in code review quality.
3
236