James Zammit

We’re excited to announce today that @mastra has raised a $22M Series A led by Spark Capital. This brings our total capital raised to $35M:

we raised a $13m seed round from 120+ of Silicon Valley’s top investors for @mastra, the leading TypeScript agent framework

Today we're launching Tasklet — an AI agent for automating your business. Unlike ChatGPT, @TaskletAI actually does the work for you: connecting to your tools, triggering automatically, and handling tasks while you sleep.

Game Arena from @generalityinc is the largest LLM strategy game tournament to date. Games are great for measuring LLMs on instruction following, long-horizon planning, and problem-solving. In fact, models that are Olympiad-level at math and coding often struggle to make accurate moves in Game Arena (median illegal-move rate: 11.4%). Game environments also resist contamination and saturation: every run is unique, so there’s no risk of training-data leakage, and the bar keeps rising as models improve. Game Arena has digitized hundreds of board games into fresh new environments, and today it’s debuting its first tournament, with models like GPT-5 High, Claude Opus 4.1, and DeepSeek V3.1 from OpenAI, Anthropic, Qwen, DeepSeek, Google, and more going head-to-head. You can watch game replays and full model reasonings on game-arena.ai. Congrats on the launch, @sanerc110 and @kaylalee278!

I'm excited to share we're shipping a new model router at @mastra today! Use 600+ models from 45 providers with a single API 🧵

@thdxr Opencontrol needs some love

in opencode 0.9.12 you can create custom tools all it takes is a file that exports a function in `.opencode/tool/mytool.ts` and it runs natively inside opencode - no mcp, no overhead

This scientist creates breathtaking visual art using chemistry.

today we're launching opencode zen zen provides the best LLMs for coding ALWAYS in the best possible configuration no more "did claude get dumber?" no more being routed to a crappy provider and we're giving it to you at cost and you can use it anywhere that enough for you?

Voice AI is growing fast, but reliability is still the #1 challenge. That’s why we built Roark - the QA + Observability layer for Voice AI. We’re live on Product Hunt today 🚀 - would mean a lot if you could upvote + comment: producthunt.com/products/roark… 🙏

LIVE5TREAM THURSDAY 10AM PT

lmfao why do we keep ending up so high on these benchmarks we haven't even tried to improve llm performance 99% of our energy has gone into packaging

@ycombinator deadline is here if you are considering applying... you should it definitely changed my life

@smthomas3 @zammitjames @dan_gauci Vibes for voice agents sounds like a disaster 🙈 The stakes are raised with voice agents for security and safety because they’re frequently deployed in healthcare admin or customer facing settings and any awkward pauses or statements are more noticeable that text chat

Voice agents built on "vibes" will fail in production (with @zammitjames and @dan_gauci from Roark) - Most teams are just "vibing" with voice agents - manually calling and testing based on feeling - Running proper simulations should be part of your CI pipeline to catch regressions and breaks - Roark allows both manual testing and scheduled runs to monitor third-party service issues - When to start using proper evals? As soon as you're onboarding customers or preparing for production - Even for simple use cases like appointment scheduling, implement evals for your main success criteria

this has been missing from american policy for a while if we are true believers in capitalism we have to be aggressive in encouraging competition way too many companies resting on what they've built up - this is anti-capitalist

The voice AI opportunity is here—but solving the last 20% is where the real challenge lies (with @zammitjames and @dan_gauci from Roark AI) - Voice AI is becoming the new frontier with companies like Roark building "datadog for voice AI" to help developers debug their agents - Popular use cases include lead generation (outbound sales calls) and appointment scheduling for industries like dental clinics and car dealerships - Getting to 80% functionality is easy, but the last 20% is where voice agents struggle significantly - Major issues: 30% of transcriptions are incorrect (some companies hire humans to manually transcribe), conversations deteriorate over time, and tool calling is notoriously inaccurate - Voice agents have trouble recovering when conversations go off-script, leading to hallucinations and poor user experiences

Some vibe pics from @ycombinator alumni demo day

@thdxr assistant-ui.com

is someone working on an embeddable LLM chat UI? the vercel chatbot template is very good - implements everything you'd need along with some cool ideas with widgets the problem is it's built on top of app router so not exactly the type of thing you can easily drop in anywhere just like we have a set of rich text editing libraries we can drop in, think we need the same for this - a lot of products will need it we're gonna end up building it unless someone is already focusing on it

Mastra (@mastra) is the open-source JavaScript framework for building agents. Companies use Mastra to automate support, build CAD diagrams, scrape the web, do medical transcription, and more. ycombinator.com/launches/Muk-m… Congrats on the launch @calcsam, @smthomas3 and @abhiaiyer!

@thdxr I just use a lambda within the same vpc that runs drizzles migrator and use pulumi's aws.lambda.Invocation to autotrigger the lambda on every deploy.

those of you with databases inside a vpc what is your migration process like? sst autodeploy can run in a vpc so there's no issue here but otherwise i deploy a lambda in a vpc that does the migration do people actually setup the custom thing for github actions?