@chinmay_bharti @amartya_jha Many congrats on the launch! Been following your story. Great progress and amazing product feature set. Sharing with my dev friends. :D

@rohanrecommends Thank you so much! Love the support and the kind words. It’s been quite a journey, and I’m excited about what’s ahead. Huge thanks for sharing it with your dev friends—means a lot! Let me know if they have any feedback or questions.

@charu_gupta9 - Thank you so much! You've been an incredible support in our Product Hunt launch. Stay awesome!

@chinmay_bharti @amartya_jha Fascinating stuff! Btw, Quick question - are you guys leveraging AST parsing for code chunck analysis before passing them through your AI pipeline. Also, are you considering LLMs on machine for your AI code reviews? Will love to know more, thanks.

@chinmay_bharti @dontdude - Thanks for the great question! Yes, we do leverage AST (Abstract Syntax Tree) parsing for code chunk analysis, and here’s how it works: 1) We connect with GitHub, Bitbucket, or GitLab. 2) We take any repository and create an AST of it, which helps us understand the entire code flow. 3) We’ve written over 30,000 deterministic checks in 30+ languages https://docs.codeant.ai/antipatt... 4) Our smart logic identifies which issues can be deterministically fixed using AST and which need AI. We can go back to the AST to check upstream and downstream functions of any call. 5) Depending on the analysis, it either goes through AI or AST. AST-based detection and fixing is almost 100% accurate. We’re already running this setup for a range of customers, from startups to unicorns. You can read more about it here - https://www.codeant.ai/customer-... We also support on-premise deployment, both on cloud and bare metal, combining AI with our rule-based engine to reduce false positives and improve code quality & security. Let me know if you’d like to dive deeper into any of this!

@sentry_co - Absolutely! Here’s a more detailed look at how CodeAnt AI handles code quality checks: 1) Rule-Based Engine & AST Parsing: We run over 30,000 deterministic checks in 30+ languages and use AST parsing to deeply understand your code structure and then pair this code flow with AI, providing insights beyond basic linting. 2) Code Complexity Analysis: We evaluate code based on cognitive and cyclomatic complexity, offering actionable suggestions to simplify complex areas for better maintainability - https://www.codeant.ai/code-comp... Dead Code Detection: CodeAnt AI identifies dead code with a confidence percentage and generates detailed reports. You can also customize scans by including or excluding specific files - https://www.codeant.ai/dead-code... Duplicate Code Detection: We detect duplicate code within and across files, providing refactoring suggestions. You can set thresholds for duplicate code detection and include or exclude specific files or folders for a more tailored analysis - https://www.codeant.ai/duplicate... Let me know if you'd like any more details!

@amartya_jha Thanks. Looking at the links now. What's AST parsing?

@dzaitzow - Thanks so much for the kind words and congrats! 🙏 To answer your questions: 1) Deeper Code Understanding with AST Parsing: We go beyond standard AI suggestions. CodeAnt AI connects with platforms like GitHub, Bitbucket, or GitLab and creates an AST (Abstract Syntax Tree) of your entire repository. This allows us to fully understand the code flow and context, enabling more accurate reviews and suggestions tailored to your specific codebase. 2) Rule-Based Engine + AI: We’ve built in over 30,000 deterministic checks across 30+ languages. Our system smartly decides which issues can be resolved using the AST alone (which ensures near-perfect accuracy) and which require AI for more complex fixes. This combination significantly reduces false positives compared to generic AI tools. 3) Security Focus: In addition to code quality, we’re deeply focused on security. CodeAnt AI scans for vulnerabilities, secrets, tokens, and more, providing security checks alongside code reviews. This is something that sets us apart from tools like GPT or Copilot. Why Trust the Reviews?: reviews are based on a combination of deterministic rule-based checks and AI, making them reliable and context-aware. The AST-based analysis ensures that our fixes are accurate and rooted in an understanding of your code’s structure, not just pattern-based guesses. We’ve already seen this benefit many teams, and we’d love to hear your thoughts once you give it a try. First 30-days is on us - https://www.codeant.ai/ai-code-r... Thanks again for your support, and let me know if you have any other questions!

@amartya_jha thanks so much for the in depth reply! Hope its helpful for other folks in the thread too!

@ashitvora - Appreciate you sharing it with your team — it’s great to hear! I’d love if you guys get started with the tool. Here’s the link for a 30-day free trial: https://www.codeant.ai/ai-code-r.... Do let me know if you need any help.

@simonas_kauzonas - Thanks a lot! Really appreciate that. I’m excited you’re going to try it out, and I’d love to hear what you think. If you or your developer friends need help or have any feedback, just give me a shout!

@jonnymiles - Thank you so much! 🙌 Our goal is to make code reviews faster and more efficient by providing comprehensive insights directly within the PR.

@runs - Thank you so much! 🙌

@vedant4510 - Thanks so much for the kind words and ongoing support! 🚀 We’re really excited to be addressing this need by combining code quality & security scanning with AI. And you’re absolutely right—this is just the beginning!

@irvin_at_subframe - Thank you so much! It’s been great having Subframe as part of the journey. Really glad to hear you’ve been impressed with the AI quality!

@kalpesh_bhalekar1 - Thank you so much! 🙌 That’s awesome to hear about your 'Aha' moment with the GitHub PR summaries—it’s one of our favorite features too! CodeAnt is a SonarQube alternative. 🚀

Heyy @jake_strack - No, CodeAnt AI doesn’t write code. We’re a code review platform that helps you find and fix code quality issues and security vulnerabilities. Our focus is on improving the health of your codebase and ensuring it’s secure!

@oliverstafurik Thank you so much! 🙌

@chinmay_bharti @utkarsh_panwar @atharva_bondre2 - Thank you so much! 🙌

@manu_chaudhary - Yes, we support self-hosted GitLab! We actually have live customers using it with their GitLab self-hosted setups. You can get started on a 30-day free trial - https://www.codeant.ai/ai-code-r...

@pelaseyed - Thanks a lot, man!

@samagra_sharma - Thank you so much! 🙌

@yash_sancheti - Yes we do scan for potential harm code or vulnerabilities, give it a try - https://www.codeant.ai/ai-code-r...

@esadik - Thanks a lot for your kind words! To answer your questions: Yes, we support multi-language code reviews—it's actually language-agnostic. As for API access, we offer it in our enterprise plan, allowing you to seamlessly integrate CodeAnt into your CI/CD pipelines for dynamic and scalable code reviews. You can get started on a 30-day free trial here - https://www.codeant.ai/ai-code-r... Let me know if you have any more questions!