Sruthi Viswanathan

@ycombinator @revnu_app @artfreebrey Go Revnu 🍀💚🤞

Revnu (@revnu_app) makes growth a solved problem. With AI now anyone can build a product, but with Revnu anyone can sell it. You never have to hire a growth person again. Technical teams focus on building, while Revnu handles the rest. Congrats on the launch, @artfreebrey & @GeorgeJeffersn! ycombinator.com/launches/QS8-r…

@ycombinator @runpentagon Let's goooo 🎉🎉🎉 @runpentagon @edgarpavlovsky

Pentagon (@runpentagon) is a coordination layer for humans and agents. Agents are everywhere at work now - coding, researching, ops, even talking to customers. But they work alone, and humans become the middleware. Pentagon is how everyone communicates, coordinates, and works as one team. Congrats on the launch, @edgarpavlovsky! ycombinator.com/launches/QS1-p…

@raidingAI @deepintxns @slashyai super thankful that i didnt dismiss it as yet another email agent 🙏

@_sruvis @deepintxns @slashyai yayyy @slashyai <3 you too

Viral YC launch week as a solo founder = balls everywhere at @deepintxns except @slashyai is in my email + slack catching them before i even know i dropped them, ops never felt so easy!!! Go Slashy 💜💚

@edgarpavlovsky @ycombinator Let's go Agentmaxxxing 🎉🎉🎉🎉

Introducing Pentagon: a coordination layer for humans and agents. ASI yields two futures: 1. AI replaces us. 2. AI empowers us to do our greatest work. We're building for the latter. Our first product is Pentagon Studio: pentagon.run An in-depth article below:

@ycombinator @andustry_hq congrats @hetdv and @JStuerken, lets goooo 🚀🚀🚀🚀

Andustry (@andustry_hq) is an AI-native broker that helps manufacturers find suppliers for industrial goods. Hardware sourcing takes months, and teams almost always overpay. Andustry helps customers save 30% on costs, while cutting sourcing time in half. Congrats on the launch, @hetdv & @JStuerken! ycombinator.com/launches/QP4-a…

@isaac_tolley17 @ycombinator @t_blom @deepintxns thank you 😊

@ycombinator @t_blom @deepintxns congrats @_sruvis !!

95% of AI pilots fail. Not because the models are bad. Because teams can't build in sync. Deep Interactions (@deepintxns) is the collaborative AI builder that ships working products in an afternoon. The future of AI isn't more prompts. It's better collaboration. Congrats on the launch, @_sruvis! ycombinator.com/launches/QOt-d…

@bilalzahalan @ycombinator @deepintxns hmmm maybe because if they collaborated in the beginning and all the ideas didnt disappear the pilot wouldnt dead and everyone wants to own it? atleast thats the case for our customers!

@ycombinator @deepintxns i'd argue most pilots fail because they're solutions chasing a problem. the collaboration friction shows up later, once nobody wants to own the dead pilot

@HoldexIo @ycombinator @deepintxns you mean our conflict resolution algorithm ;) not sharing that, thank you 😊

@ycombinator @deepintxns Deep, tech was never the real challenge. Getting a room full of brilliant minds to stop talking and actually build in sync is. How does @deepintxns prevent 'prompt conflicts' when everyone wants to steer the AI agent? @_sruvis congrats for the launch btw 🎉

@emax Thanks Maxxxxx 😍😍

Excited for the brilliant team that's working on this > 👏

@orangecollectv @AquaShieldai @deepintxns thanks for catching us just launched x.com/ycombinator/st…

Deep Interactions (@deepintxns) - Collaborative AI app builder for teams Deep Interactions uses humans-in-the-loop to generate clean, maintainable code that typical AI builders can't match. their context layer keeps teams in sync while reducing token usage and development time. Founded by @_sruvis deepinteractions.ai

today's launches span AI-powered building protection, collaborative vibe coding, and streamlined regulatory compliance for life sciences. here's the 3 newest YC P26 startups announced in the last 24 hrs 👇

@tpersikova @deepintxns lets goooo champion @tpersikova

And boom, we're YC live! If you've ever burned through your token limit with nothing to show for it, and had to scour across 20 Slack threads looking for context for your vibecoding session—this little spider is here to help 🕷️ 😊 Give @deepintxns and @_sruvis a follow, share with your feed. We're just getting started!

@ctrlswftdelete @ycombinator @deepintxns 100%

@ycombinator @deepintxns the temptation to think AI is a solo race is the real problem; collaboration tools might be the UX fix AI never had, which is honestly the kind of detail YC should hype more

@MaksimXBT @ycombinator @deepintxns glad you asked > forbes.com/sites/jasonsny…

@ycombinator @deepintxns 95% fail rate sounds fixed, but what happens when the team is in sync and the model is still bad

@deepintxns is live on YC today 🥳🥳🥳

@deepintxns is now live 😊😊😊🥳

How should UIs evolve in the Age of AI? Everyone's declaring GUIs dead. But we're still living through the command-line era of AI — typing instructions into blank text boxes and hoping for the best. open.substack.com/pub/personalfu…

Introducing ml-intern, the agent that just automated the post-training team @huggingface It's an open-source implementation of the real research loop that our ML researchers do every day. You give it a prompt, it researches papers, goes through citations, implements ideas in GPU sandboxes, iterates and builds deeply research-backed models for any use case. All built on the Hugging Face ecosystem. It can pull off crazy things: We made it train the best model for scientific reasoning. It went through citations from the official benchmark paper. Found OpenScience and NemoTron-CrossThink, added 7 difficulty-filtered dataset variants from ARC/SciQ/MMLU, and ran 12 SFT runs on Qwen3-1.7B. This pushed the score 10% → 32% on GPQA in under 10h. Claude Code's best: 22.99%. In healthcare settings it inspected available datasets, concluded they were too low quality, and wrote a script to generate 1100 synthetic data points from scratch for emergencies, hedging, multilingual etc. Then upsampled 50x for training. Beat Codex on HealthBench by 60%. For competitive mathematics, it wrote a full GRPO script, launched training with A100 GPUs on hf.co/spaces, watched rewards claim and then collapse, and ran ablations until it succeeded. All fully backed by papers, autonomously. How it works? ml-intern makes full use of the HF ecosystem: - finds papers on arxiv and hf.co/papers, reads them fully, walks citation graphs, pulls datasets referenced in methodology sections and on hf.co/datasets - browses the Hub, reads recent docs, inspects datasets and reformats them before training so it doesn't waste GPU hours on bad data - launches training jobs on HF Jobs if no local GPUs are available, monitors runs, reads its own eval outputs, diagnoses failures, retrains ml-intern deeply embodies how researchers work and think. It knows how data should look like and what good models feel like. Releasing it today as a CLI and a web app you can use from your phone/desktop. CLI: github.com/huggingface/ml… Web + mobile: huggingface.co/spaces/smolage… And the best part? We also provisioned 1k$ GPU resources and Anthropic credits for the quickest among you to use.

@ycombinator @Silmarildev @aumup001 Congratulations 🎉🎉🎉

Silmaril (@Silmarildev) is the first self-healing prompt injection defense. It catches 2x more attacks 10x faster than leading defenses, and retrains continuously to protect your full AI stack, including agents like Claude Code and OpenClaw. Congrats on the launch, @aumup001 and @EduardoVel36291! ycombinator.com/launches/Pvl-s…