frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Launch HN: Agnost AI (YC S26) – Extract user feedback from agent conversations

https://agnost.ai
16•laalshaitaan•1h ago
Hey HN, we’re Shubham & Parth, childhood friends building Agnost AI (https://agnost.ai), product analytics for teams building chat and voice agents.

We read production conversations and find behavioral failures like users rageprompting (cursing at the agent), repeatedly rephrasing the same request, correcting the agent, asking for missing features, or leaving after an answer that was technically successful.

We have an interactive demo with no signup here: https://app.agnost.ai?demo=true

Here's a demo video: https://www.tella.tv/video/agnost-ai-launch-hn-demo-9haa

The core problem is that chat and voice products do not have the same metrics as web apps. When the product interface is language, clicks and funnels become much less useful. Users also rarely give explicit feedback, and when they do it's usually sugarcoated. I barely type /feedback in Claude or Codex myself. Most users just curse, ask again, correct the agent, or leave. So product engineers get technical visibility from latency, errors, and traces, but still have to guess whether users got what they wanted.

We got here after building around agents for the last year and got a couple of founders asking for something like a PostHog for conversations for the AI assistants they were building.

We are not trying to be in the observability or evals space. Observability tells you what happened technically. Evals validate cases you already know. We're more on the discovery side like what users wanted, where they got frustrated, what they asked for repeatedly, and what new evals should exist.

Teams send us agent conversation messages through SDKs or OTel, optionally with metadata like account, plan, source, organization, etc. We cluster conversations into product-specific intents. Feature requests and bugs are default categories; most other clusters are created dynamically from the customer’s data and evolve over time. You can create your own cluster in plain English. If a cluster gets too broad, we split it. If a new pattern appears, we suggest it.

One AI video editor company used Agnost AI to find feature requests hidden inside chat. The biggest one was that around 70 users wanted auto-subtitles, but users said it as “add this text in this frame” 12x in a single session, “can you caption it”, “give me transcript of audio” and variations across languages. The team later built the feature.

Doing this over millions of messages without sending everything to an LLM was the hard part initially. In ClickHouse, “fetch the last 50 events by time across conversations” and “fetch all events in this conversation” want different sort orders, so we had to iterate a lot on sorting keys, partitions, materialized views, and projections.

For finding new clusters, sending everything through an LLM was too slow and expensive. HDBSCAN-style embedding clustering also gets painful at scale because of pairwise comparisons. We first split conversations into segments based on cosine drift, run BIRCH to compress the candidate space, and then use HDBSCAN-like clustering on the smaller set. For matching existing clusters, we use embeddings, smaller classifiers/BERT-style models, and LLMs only as fallback for ambiguous cases.

We’re live with multiple companies and ingesting ~1M chat and voice messages per day. Pricing is public: Starter is free, Pro is $499/month, and Enterprise is for higher volume, security, retention needs. We use each customer’s data only for that customer. We are SOC 2 Type 1 compliant, Type 2 is in progress, and our SDKs are on PyPI and npm.

We’d love feedback from the HN community and people building chat or voice agents: how do you detect these signals today, what feedback methods have worked, and what would block you from trying this? Happy to answer questions and take criticism.

Comments

zuzululu•24m ago
why would i pay $499/month for this when codex costs $199/month and can do everything you described
laalshaitaan•4m ago
codex is great for like a one-time/overview analysis on a handful of transcripts. we usually serve to companies where the volume is >10k messages & continuous ingestions + with claude/codex it messed up this + metadata linking of the user like what plan are they on, when is it expiring, etc.

although we had a few customers who come to us after running this for a while so at smaller volume it does work well.

m_kos•20m ago
> Rageprompting

Lovely name! I implemented profanity monitoring in my Hermes setup to identify "learning opportunities" for my agents. It is quite useful. If you are budget-conscious, one challenge is determining what is the smallest number of previous rounds that Hermes needs to correctly infer what it did wrong. Curiously, Claude Code is horrible at figuring out what it did wrong. I often read its memories, and they are rarely useful.

laalshaitaan•12m ago
haha yea, i even got the domain rageprompt dot com like a couple of days ago lol i love the name too.

for profanity, did you define keywords or just let the agent figure out rage stuff?

how many rounds did you set for the hermes? claude doesnt work yea on its own, one of my friends set us up for their claude lol

Measuring Input Latency on Linux: X11 vs. Wayland, VRR, and DXVK

https://marco-nett.de/blog/measuring-input-latency-on-linux-x11-vs-wayland-vrr-dxvk/
118•hoechst•1h ago•41 comments

The Tower Keeps Rising

https://lucumr.pocoo.org/2026/7/13/the-tower-keeps-rising/
34•cdrnsf•52m ago•9 comments

Your 'app' could have been a webpage (so I fixed it for you)

https://danq.me/2026/07/09/your-app-could-have-been-a-webpage/
432•MrVandemar•3d ago•299 comments

Show HN: Opening lines of famous literary works

https://www.verbaprima.com/
92•plicerin•2h ago•58 comments

Are we offloading too much of our thinking to AI?

https://www.artfish.ai/p/offloading-thinking-to-ai
195•yenniejun111•2h ago•175 comments

S&P downgrades Oracle to BBB – only one notch above junk level

https://www.heise.de/en/news/S-P-downgrades-Oracle-to-BBB-only-one-notch-above-junk-level-1136347...
74•gepeto42•53m ago•49 comments

How to stop Claude from saying load-bearing

https://jola.dev/posts/how-to-stop-claude-from-saying-load-bearing
187•shintoist•6h ago•283 comments

Kontigo (YC S24) Is Hiring (Head of Security)

https://www.ycombinator.com/companies/kontigo/jobs/uNttrlv-head-of-security
1•jecastillof•50m ago

Launch HN: Agnost AI (YC S26) – Extract user feedback from agent conversations

https://agnost.ai
16•laalshaitaan•1h ago•4 comments

A tiny cell that broke a big rule of biology

https://grist.org/science/nitrogen-cycle-cell-discovery-nitroplast-science-fertilizer-algae-bacte...
63•gumby•5d ago•10 comments

Beautiful Type Erasure with C++26 Reflection

https://ryanjk5.github.io/posts/rjk-duck/
83•RyanJK5•5h ago•33 comments

Paxos Made Simple (2001)[pdf]

https://lamport.azurewebsites.net/pubs/paxos-simple.pdf
44•grep_it•4d ago•3 comments

The zero-cost fallacy: open-source software in the agentic era

https://www.thoughtworks.com/insights/blog/open-source/zero-cost-fallacy-open-source-agentic-era
8•backlit4034•3d ago•0 comments

Show HN: I RL-trained an agent that trains models with RL (for –$1.3k)

https://github.com/Danau5tin/ai-trains-ai
74•Danau5tin•5h ago•31 comments

Show HN: Juggler – an open-source GUI coding agent, by the creator of JUCE

https://github.com/juggler-ai/juggler
97•julesrms•1d ago•53 comments

Punch yourself in the face with reality

https://adi.bio/reality
130•AdityaAnand1•6h ago•63 comments

Superoptimizer – A Look at the Smallest Program

https://dl.acm.org/doi/epdf/10.1145/36177.36194
20•linggen•3d ago•4 comments

How the FSF sysadmins block botnets with reaction

https://www.fsf.org/blogs/community/blocking-botnets-with-reaction
118•pseudolus•2d ago•36 comments

European "age verification" "app" forcing everyone to use Android or iOS

https://github.com/eu-digital-identity-wallet/av-doc-technical-specification/discussions/19
293•roundabout-host•9h ago•181 comments

The Agentic Loop: Three loops in a trench coat

https://www.bobbytables.io/p/the-agentic-loop-three-loops-in-a
16•btables•3h ago•3 comments

Australian energy retailers must provide three hours of free daytime electricity

https://lenergy.com.au/free-daytime-electricity-is-coming-heres-how-it-actually-works/
227•i2oc•13h ago•308 comments

Alternative(s) to run CUDA on non-Nvidia hardware

https://www.hpcwire.com/2026/07/09/spectral-compute-aims-to-set-cuda-free-will-it-succeed/
111•alok-g•9h ago•62 comments

A metallurgist's doubts about self-replicating probes

https://www.centauri-dreams.org/2026/07/10/a-metallurgists-doubts-about-self-replicating-probes/
118•EA-3167•1d ago•38 comments

Our Amish Language

https://www.thedial.world/articles/news/amish-pennsylvania-dutch
78•NaOH•15h ago•59 comments

Indian scientists produce most detailed 3D atlas of the human brainstem

https://www.bbc.com/news/articles/cg53l737v1qo
165•BaudouinVH•11h ago•20 comments

Germany set to restrict its Freedom of Information Act

https://www.dw.com/en/germany-freedom-of-information-act/a-77939695
228•robtherobber•6h ago•141 comments

The git history command

https://lalitm.com/post/git-history/
403•turbocon•16h ago•285 comments

No Spanish reading crisis?

https://www.commonreader.co.uk/p/no-spanish-reading-crisis
56•jruohonen•6h ago•93 comments

Tensor Is the Might

https://zserge.com/posts/tensor/
36•eatonphil•4h ago•20 comments

Satellite Tracker – Live Map of Starlink and 30k Satellites

https://satellitemap.space/
146•rolph•15h ago•73 comments