im trying Appwrite , Keywords AI (YC W24) , tavily , mem0 for the first time. wish me good luck))

LLM Evals is now part of every conversation in Silicon Valley. At Keywords AI (YC W24), we process over 500M+ LLM logs monthly. We see that many developers struggle with reducing hallucinations, making agents follow instructions, and retrieving memory quickly and correctly. Want tips on

Cheers to RAG & LLM Evals — Talk & Beer Social in SF Join Keywords AI (YC W24) x Ragie for a chill Talk & Beer Social at Olfactory Brewing. 🗓️ June 26th | 🕕 6–9 PM 📍 Olfactory Brewing & Blending, 2245 3rd St, San Francisco 👉 RSVP here: lu.ma/zdtzmg3j

We’re hosting a chill happy hour just 5 min from the YC HQ this Thursday - AI builders and YC founders welcomed! 🍻 Free craft beer & pizza 📍 Olfactory Brewing & Blending, 2245 3rd St 🗓 June 26, 6PM Come hang!

IMPROVEMENT: Better UI for choosing LLMs for the AI gateway We improved our Models page to help developers choose LLMs for their AI apps more easily. You can now filter models by provider and see which ones have API keys added.

7 am, last Sunday. Walked out and saw the team already up, watching football together. We’ve all been living under the same roof for a while, and moments like this remind me how much our lives overlap — not just building and working, but actually sharing everyday life. It’s

We just shipped support for image inputs in Experiments. More AI apps are using images alongside text — especially agents like QA bots, browser agents, or travel assistants. But most prompt testing tools only support text. If your dataset includes images (e.g. product photos,

IMPROVEMENT: Custom date range for LLM log filtering We added a custom date range option for filtering LLM logs. Now you can select any specific time period to view and export the exact logs you need. Try it out today!

I used to evaluate my work by how well I planned the week or month ahead. Now, I evaluate it by how well I execute. Are the must-do tasks actually done? Did we make progress today? When I first started building Keywords AI (YC W24), I was obsessed with planning. I’d spend entire days

Understand the AI Agent Stack (Frank’s Notes - Day 1) As a DevRel, I’ve spent the last few months diving deep into the AI agent ecosystem — trying to map out what’s real, what’s useful, and what’s just noise.

Brendan Falk Keywords AI (YC W24)

Integrating Keywords AI into AI Agentic Apps 🧵 Give Your AI Agents Enterprise-Grade Observability with Keywords AI Taking a closer look at this open-source project Eliza - AI Agent Orchestration (Wei): github.com/Anora-Labs/eli… Building for 100 Agents Hackathon. Ariel Weinberger

Prompt engineering is being replaced by context engineering. Prompt engineering started as a way to help LLMs better understand human needs. It worked well — especially for early AI apps that were mostly wrappers around ChatGPT or simple workflows. But things are changing fast.

Retell x OpenAI 🤝 OpenAI just spotlighted how we’re using GPT‑4o to automate real phone conversations—no scripts, no code, no hold music. 📉 80% cost reduction 🌟 90+ NPS 🚀 Live in days, not weeks Read the full openai.com/index/retell-a…

Our first party Keywords AI (YC W24) happened yesterday with Ragie. Huge thanks to everyone who came out - excited to keep building with this community. Also... our sales guy might’ve had a bit too much beer. If anyone finds him, please return him to the office. 😅

Built my own AI scheduler using the OpenAI JS Agent SDK — now with full tracing via Keywords AI (YC W24). Take a look👇