Kritin Vongthongsri

building evals @deepeval @confident_ai

[Jeffrey 🐬 confident-ai.com @jeffr_yyy]

Today, we're proud to announce DeepEval 4.0 — the AI evaluation harness for vibe coding agents. Our biggest and boldest release yet.

A long thread 🧵 @deepeval :

[Jeffrey 🐬 confident-ai.com @jeffr_yyy]

Vibe coding is dead. Vibe coding without vibing is the way forward.

Coding tools like @Claude code, @cursor_ai, and more don't make real progress in improving agents.

The last mile is what's difficult.

[Kritin Vongthongsri @kritinv07]

what do people even do when waiting for Opus 4.7 extra high on cursor?

[Brian Neville-O'Neill @bnevilleoneill]

I’m running LLM eval office hours today with @confident_ai 🧪

If you’re building anything with AI, drop a prompt + model output, and I’ll show where it breaks.

I’ll look at:
correctness
completeness
where it might fail in real use.

Just quick, specific feedback
#ai #LLM

[DeepEval @deepeval]

My sister just got released, DeepTeam v1.0, 100% open-source, Apache 2.0 red teaming for LLMs.

⭐ Star on GitHub to stay on top of the latest developments in AI security and safety: https://t.co/FwtIQ5xiSA

[Kritin Vongthongsri @kritinv07]

What makes platform UI enterprise ready?

[Kritin Vongthongsri @kritinv07]

@oleg_golev 😂

[Kritin Vongthongsri @kritinv07]

Making it so easy to view and evaluate threads/conversations on @confident_ai.

[Kritin Vongthongsri @kritinv07]

Turning to https://t.co/l5mJsGtRvq because https://t.co/JTjjHlfS3g is gone for good 😢.

[Kritin Vongthongsri @kritinv07]

We're cooking up 👨‍🍳 something for our @confident_ai users...

[Kritin Vongthongsri @kritinv07]

Most people run single-turn evals on chatbots.

But that’s not enough.

Conversations aren’t Q&A — they happen over multiple turns.

This means your chatbot must stay context-aware across the dialogue, not just accurate in isolated responses.

@deepeval, we’ve seen too many teams evaluate chatbots the wrong way.

So, we wrote a comprehensive guide on how to evaluate all chatbots properly, end-to-end.👇

🔗 https://t.co/L0iGyIct9X

[Jeffrey 🐬 confident-ai.com @jeffr_yyy]

who took down pypi downloads? https://t.co/ohpv3WZjVy

[Kritin Vongthongsri @kritinv07]

This is why you shouldn't vibe evals.

[Santiago @svpino]

Vibe-coding feels like magic.

Until you're the one cleaning up the magic later.

[Kritin Vongthongsri @kritinv07]

@LTantichot 🪞

[Kritin Vongthongsri @kritinv07]

What’s more important in growing oss: building or writing docs?

[Kritin Vongthongsri @kritinv07]

@LTantichot @greptile @soohoonchoi @helicone_ai @coleywoleyyy @0xpunnk @confident_ai @autumnpricing @ay_ushr @helixdb @xav_db @tryrevyl @anamhira @sixtyfourai it's ok to tell your friends you love them

[Kritin Vongthongsri @kritinv07]

@LTantichot @spicy_liu @RajitWrites I didnt know u could do this

[Jeffrey 🐬 confident-ai.com @jeffr_yyy]

ten kay stars @deepeval

[Kritin Vongthongsri @kritinv07]

At @confident_ai, we’re focused on making evals great.

But since we love our users very much, we’ve also just 5×’d the tracing analytics on our platform.

Now you can:

🔍 Trace analytics — follow every request end-to-end
⏱️ Span analytics — see latency and cost per component
📊 Model analytics — compare performance, latency, and cost across models
👥 User analytics — understand usage patterns and behavior
⚠️ Error analytics — track and reduce failures over time

[Jeffrey 🐬 confident-ai.com @jeffr_yyy]

We quietly released a new OS package a few months ago. It has now over 600 stars.

https://t.co/2nWrxgfZGd