Subscribe
Sign in
Jolt AI — AI assistant for 100k to multi-million line codebases
Jolt is your instant codebase expert. It's an AI codegen and chat tool for 100k to multi-million line codebases that automatically identifies context files, handles multi-file changes, and matches your code style.
55
Replies
Best
flo merian
Hunter
📌
AI is shaping how we code.
There are amazing products for rapid prototyping and small codebases. However, larger, real-world codebases are still a challenge for most of them.
Jolt AI solves it with its codegen and chat tool — purpose-built for production codebases and perfect for engineering teams with 100K to 1M+ line codebases.
May this be the future of production-ready AI developer tools. Keep up your great work, @yev_yev_yev and team 👏👏
Report
Share
Hunter
@fmerian Thanks so much for hunting and posting Flo! We're excited to get Jolt in the hands of every developer ⚡️
Hey Product Hunt - thanks you all so much for the support!
We built Jolt to address a major gap in AI coding tools - they struggle with larger, real-world codebases and cannot accurately determine the context files for your prompt.
We developed a novel way for Jolt to understand large codebases and automatically determine context files. Our approach scales to multi-million line codebases - the largest codebase using us in production is over 8M lines.
Manually selecting context files is a non-starter when you're working on larger codebases or unfamiliar code. It's a broken product experience. Most AI coding tools rely on some flavor of vector-embedding RAG to determine the files related to your prompt. The reality is that vector-embedding search is not effective on code, and there is a sharp drop in efficacy as codebase size increases.
Here's what folks are using Jolt for:
- Writing code, tests, and refactoring
- Onboarding developers
- Brainstorming a feature's implementation
- Asking questions about OSS repos
- Writing documentation, including mermaid charts
- Contributing across the stack
Thanks again and let us know what you think
@yev_yev_yev Looks great! How does this compare to Cursor? I've been painstakingly struggling to get good results from Cursor when I combine searches across 5 code repos. Would be curious to see what you guys do differently
@kevinyun Great question. We developed a new way for AI to understand large codebases - so even if your repo is millions of lines, Jolt can accurately identify the context files. With Cursor and others, you'll likely need to manually select the context files - that is very tedious and time consuming.
Jolt also makes file changes on its own, without you pointing the cursor to tell the AI where to insert code. Happy to give a demo and share more ⚡️
@seungwhan Thank you - let me know if you'd like a demo or have any questions
@sentry_co Great question - Jolt automatically figures out the best model to use so you don't have worry about. It all works seamlessly behind the scenes
@yev_yev_yev I always wondered about how a switching mechanism like that would work. Like a LLM for figuring out which LLM to use for a particular prompt. Could you elaborate how yours work?
@sentry_co You start to see what types of tasks certain LLMs are best at and then you establish criteria for selecting one model over than based on the task or question at hand
Hunter
@astrodevil thanks for the support, Anand! please help us spread the word on X 🙏
Wow, unreal that it actually matches the codestyle too - Love finding new dev tools to aid myself & the team with out work.
Top job on hunting this Flo, I think you've smashed it :)
Hope you get #1 today, best of luck Yev, Carlos & Jon!
Hunter
@cranqnow thank you, Sam -- fingers crossed! would you help us spread the word on X? repost this
@yansoul Thank you - our goal is to help developers understand code, even in unfamiliar parts of the codebase
Hunter
@mikita_aliaksandrovich thank you, Mikita! would you help us spread the word on X? enjoy your upcoming launch ✌️
Day one with Jolt went smoothly, it really goes above and beyond with code updates. Keep up the great work!
Incredible! Going to start using this with my company soon as the ability to work with large codebases has always been a limitation of other tools.
@danielmason1 Thanks so much Daniel! Excited to help your team ship faster
@iman_aqas1 Thank you Iman! Our goal is for Jolt's code to look like your code!
@abrams Thank you. IMO this is one of the most important problems to solve in codegen today
Great to see a product with effort on big codebase! Exciting. Congrats on the launch! 😀
Hunter
@sacha_dumay spot on, Sacha! this is definitely a challenge.
please help us spread the word on X, let's push production-grade codegen!
Been following the Jolt team for some time, favorite new AI coding tool for large codebases!
Hunter
@oreznichky1 your words just made my day, Simon -- please help us spread the word on X!
I recommend it because it makes it much easier to work with large codebases. With this tool, you can easily work with projects from 100 thousand to several million lines of code. It instantly detects the context of files, helps with changes to multiple files at once, and adapts perfectly to the code style.
I am pleasantly surprised at how fast the AI coder space is improving.
You are solving a genuine problem. Lovable and bolt etc are good for single page apps but real world tools require multiple modules and at times hundreds of files.
Congratulations on the launch and best of luck to the entire team