Rahat Kabir
@rahat_neuron
Followers
23
Following
3K
Media
59
Statuses
309
AI Engineer | Obsessed with the potential of LLMs to address real-world challenges. Building for need, not just novelty.
Joined February 2020
Day 4 🚀 my productivity app with ai analysts almost done, 2 users are testing this > added a feature like streak > polished setting page > notification sound > >feel like lightweight enough for MVP. i'll test it with at least 15 users before shipping. happy Building
2
0
1
Day 3 of building my AI-integrated productivity Pomodoro app.. 🚀 when I add a new feature, I check first and cli. cli's first approach is better. added AI in my app. Engineered to reduce token. Happy building 🚀
1
0
1
I'm using @openclaw so hard. > learning linux command with real testing. > I ask my assistant, it teach me those command and apply it’s own OS. > it show me real output of those command. this assistant can run all of command on It's own OS.
2
0
2
I'm running @openclaw using GLM model. pretty good. maybe i should use Opus4.5( for it's personality) what model do you guys are using?
1
0
1
I'm building a prompt manager. When I read Karpathy's agentic workflows or Peter's insights, I saved them instantly. Now when I'm coding and need a reference, I just find it. I can import/export prompt and built it test prompt. Tech : Nextjs,Shadcn,postgress,fastapi,docker
1
0
4
Day 2 🚀 CLI testing is the best way to validate project features. Tested 46 CLI cases already 🥹 Building an AI-powered Pomodoro assistant Adding unique features to help people track time easily & stay productive. #BuildInPublic
2
0
5
Yes, I’m building another Pomodoro timer. 😅 Meet PomoAI: A timer + AI Data Analyst. Instead of guessing, I ask: "What work i do most this week?" The Stack: 📱Flutter & Riverpod 🗄️ SQLite (Local-first) LLM MVP Home screen 👇 #BuildInPublic
1
0
5
If you are using clawdbot/moltbot/openclaw. How are you using it? Give local machine access? VPS? VM? AWS or anything?
0
0
0
Try it https://t.co/ckUmccWWH2
I just designed a landing page for this product. I'll iterate regularly. I'm going to deploy it soon.
0
0
0
Sorry Claude, you gotta commit more if you want the top spot 😉
0
0
0
Managing multiple AI provider APIs is a nightmare. @huggingface Inference Providers fixes this. one API → hundreds of models from Cohere, Groq, Replicate, Together AI, and more. text generation, image creation, embeddings, all in one place. starts free, no vendor lock-in.
4
4
18
I just designed a landing page for this product. I'll iterate regularly. I'm going to deploy it soon.
I'm working on a project to solve a simple problem. we save prompts everywhere and then forget where. This tool lets you store, search, edit, delete, and test prompts instantly inside the app. Teck stack : FastAPI,NextJs,Clerk,Postgresql for DB.
0
0
0
I'm working on a project to solve a simple problem. we save prompts everywhere and then forget where. This tool lets you store, search, edit, delete, and test prompts instantly inside the app. Teck stack : FastAPI,NextJs,Clerk,Postgresql for DB.
3
0
4
LLMs are growing fast, but prompt management hasn’t kept up. Save prompts randomly → can’t find them later → wasted time. I’m building a simple app to fix this. You can run the prompt inside the app. Tech stack : FastAPI, NextJS,shadcn, Postgres.
2
0
1
Don’t deploy services across multiple AWS regions unless you really need to. It gets messy, costs add up fast, and later you won’t even remember where things are running 😅 I now stick to 1–2 regions (like Mumbai/Singapore). Life is much easier. 🚀
0
0
1
Solving turing machine for exam - But Urge to build something
0
0
2
So basically we can watch everything Claude Code does by setting a few env variables to connect it to LangSmith. See all the prompts, tool calls, and decisions Claude Code makes while coding. Super useful for debugging.
Claude Code 🤝 LangSmith Curious what Claude Code is doing behind the scenes? Or want observability into critical workflows that you’ve set up with Claude Code. With our new Claude Code → LangSmith integration, you can view every 🤖 LLM call and 🔧 tool call Claude Code makes.
0
0
2
How good are OpenAI's skills for Codex? Has anyone tested it? https://t.co/oeBz8MKU77
github.com
Skills Catalog for Codex. Contribute to openai/skills development by creating an account on GitHub.
0
0
1
Why did you fine-tune or prune any model and test it on eval last time? Me at August 10
0
0
1