Emre Ozsahin

Just shipped RagPack v1.0 🚀 RagPack is an open source lightweight RAG infrastructure for developers that want a cost efficient solution. Designed with Ollama first experience, works with OpenAI as well. ragpack.dev #buildinpublic #opensource #rag #ai #ollama

@aslambeg84 Happy to connect Aslam

can we make 450! if u r in tech say hello 👋 let's connect

@joseph_mugen builders 💪💪

Reddit is full of bots TikTok is full of brain rot LinkedIn is full of AI slop Instagram is full of softcore p*rn What is X full of?

@Liam_McDaniel3 Hey Liam, happy to connect!

Hey Yall, Lets make new connections Looking to connect with people into: 🤖 SaaS 📷 Startup 📷 Building in public Say hi and let’s connect!

@the_real_ori @kinmorychin In memory embeddings is a different decision here, you would need to re-embed all the chunks after each build or restarts. How many documents do you typically have and how big are they?

@kinmorychin Lightweight RAG is the right call at this stage. Full vector infra for a few hundred docs is procrastination. What does the agent do with no good answer, hand off or admit it?

Every open-source RAG project I tried was too heavy to ship — a vector DB, an orchestration layer, piles of abstractions I didn't need. So I wrote my own minimal one instead. It's the 3rd agent I've scaffolded: a customer support agent for Kinmory. Chat widget + lightweight RAG + a ticketing system. My RAG is basically chunk → embed → cosine search over in-memory embeddings. No extra infra, cheap to run, fast enough at this scale. Shipping it to the Kinmory site tomorrow. Then the long game: continuously improving the agent's answer quality.

@SidDegen @shedntcare_ Hmm no docker part is interesting definitely. Surely it would cause issues when you try to deploy to different environments no? I wonder how are they solving that problem. You can always self-host lightweight RAG infras btw.

the interesting part isn't that alibaba shipped another vector db — it's that there's no db to ship. `pip install zvec` and it runs in-process, no server, no docker. @shedntcare_ flagged it, and the rent-a-vector-db model just got a real question mark 🧵

@16vchq ragpack.dev

drop your startup link get some traffic

@maxzrco congrats that's huge!

I reached $3k MRR. Feels unreal. Still dealing with imposter syndrome and wondering if it'll last. But I'm going to keep building every single day. $5k MRR is next.

@vikrambuilds both :)

Is AI making us more productive or dumb?

@rihntv It's tough for sure. The expectations have risen as much as the tools have progressed. But you can still use them effectively to make a difference.

Horrific time to be an entry level job seeker

@jfkcyrus Hi Jirka, happy to connect!

Hey! Looking to connect with people building in: 💻 Devs 🍽️ SaaS 🚀 Tech 🔥 Web App 📱 Product Development And more ofc Say hi and let's connect👋

@Surendar__05 Yes because it's still a good differentiator.

In the AI era, are you still grinding on LeetCode in 2026?

Heads up if you're on Vercel's free tier. Recently got an email about reaching my free usage quota, one route was burning ~3 min of Active CPU per billing window. Fix: server-side caching with Next.js's unstable_cache. More below.. dev.to/emrecodes/cutt…

@gregisenberg Great time to be a developer!

The computer is being reinvented in the agentic era: - The model is the new CPU. - The harness is the new OS. - Hallucinations are the new bugs. - The context window is the new RAM. - Skills are the new apps. - Markdown files are the new config. - Evals are the new QA. - Context is the new moat. - Permissions are the new firewall - Trust is the new bottleneck. - Prompt is the new programming language - Agent is the new software. Anything you dream of, you can build. This is the greatest time ever to be building with computers.

@forgebitz Easy work didn't disappear. The bar for what counts as easy just moved up. The "hard" problems of today will be tomorrow's easy work too.

software engineering is now draining because all the "easy" work is done you used to be able to do productive but relatively simple work some days (writing some new endpoints, making some dashboards) now all that is left is the hard part

@ThePeterMick Extremely hard

Startup marketing is hard ○ True ○ False

@anupamrjp Reviewing the occasional garbage it spits out :)

Hey founders,If AI is handling most of your coding… What are you spending your time on now?

@MnFounder That's huge congrats!

Good night, builders. Today was a big one. 1,100 people are now following this journey, and I’m honestly grateful. I also wrote about why side projects matter when layoffs happen. Now it is time to rest and keep building tomorrow.

@PashaBorsai Let me know if you figure it out 🤣

What's the fastest way to get your first 10 paying users?

@Monika2r Depends on what you are trying to build 😎

Which database is best ??

@_Chandan_17 You get more power for you buck assuming you have a monitor. You can do an openclaw/ollama setup as well. You can even run it as a local server since it's silent and consumes low power too.

@EmreCodes1 And why is it more important to you than macbook??

Developers, if you have $2K. What are you buying ? -Mac mini M4 -MacBook Air M5