Trying to find more like minded people on X who are actually building.
If you’re into:
🤖 AI/Agents/LLMs
⚙️ Backend/Python
🗄️ Databases
🧠 RAG/GenAI
🚀 SaaS/Startups
🔧 Open Source
I want to see what you’re working on.
Drop your project or what you’re building below 👇
🚨 SEND FREE DIWALI GIFTS TO YOUR LOVED ONES TODAY 🚨
order anything for your friends or family through @iradotso today, and it's on me!
here's how to claim it:
1. place an order through ira.
2. take a screenshot of your order confirmation.
3. once the gift is delivered, share both screenshots (order confirmation + delivery) by reposting and commenting on this post.
that's it. we'll reimburse you within 5–7 days!
just fyi —
@iradotso is invite-only. if you don't have an invite but want to participate, drop a comment below and our team will share one with you 👇🏻
the fine print:
a. valid today only (november 8)
b. maximum reimbursement: ₹500 per person
c. reimbursement within 5–7 days
@kasike So i work with Lucknow AI labs and they have good connections with news channels and all so from there we get awadhi audio and for text there are multiple ways like researchers around awadhi, newspapers publications and more.
Are you from lucknow Or nearby and how you know awdh
@b4dgerr It depends how you use it but for a moderate to high level coding and rest tasks, you have enough of tokens.
Though my roommate actually eat all the tokens every month even after getting lots of reset because he actually makes 5-6 projects in parallel🙂
Trying to find more like minded people on X who are actually building.
If you’re into:
🤖 AI/Agents/LLMs
⚙️ Backend/Python
🗄️ Databases
🧠 RAG/GenAI
🚀 SaaS/Startups
🔧 Open Source
I want to see what you’re working on.
Drop your project or what you’re building below 👇
PostgreSQL looks boring until your app gets real traffic.
Indexes. Transactions. Connection pools. Query plans. Locks. WAL.
None of these are “AI”.
But they can matter more than your prompt.
AI engineering still needs good backend engineering.
@Abisuga_ Appreciate it. I’m getting deeper into the backend side of AI agents, so I’m especially interested in this modular approach. What are you currently building?
DeepSeek Harness is an interesting direction for AI agents.
It treats models, tools, sandboxes, sessions, scheduling and UI as replaceable modules.
That feels closer to good backend design.
𝗠𝗼𝗱𝘂𝗹𝗮𝗿𝗶𝘁𝘆 beats rebuilding the stack whenever one component changes.
Cloudflare made AI coding agents more interesting.
Cursor Cloud Agents can now run inside Cloudflare-controlled sandboxes.
The model plans. The sandbox executes.
Agents are becoming infrastructure problems, not just model problems.
@ABG__UB Yeah, that makes much more sense. Secure-by-default with explicit overrides is the right approach for agents. I’m working on similar AI/backend problems, would be good to stay connected and compare notes as we build.