Home›Companies›Mastra

Mastra

The Javascript framework for building AI agents, from the Gatsby devs

Mastra is the easiest way to build a production AI application, built by the team that brought you Gatsby. We provide a complete set of AI primitives: workflows for complex operations, agents for autonomous decisions, RAG for knowledge integration, and evals for quality & accuracy. Mastra is open-source, modular, and built in Typescript for product developers. You can deploy Mastra on any cloud, and it gives you a straightforward syntax to solve the hardest problems building with AI: human-in-the-loop workflows, agent memory and tool selection, and output quality.
Mastra
Founded:2024
Team Size:8
Status:
Active
Group Partner:Gustaf Alstromer
Active Founders

Sam Bhagwat, Founder

Founder/CEO of Mastra. Spent last ~8yrs as cofounder of Gatsby.js, an open-source React website framework, scaled to $5M ARR, sold to Netlify. Eng <20 at Zenefits (W13) and Plangrid (W12). Spent two years knocking doors. Stanford '11.
Sam Bhagwat
Sam Bhagwat
Mastra

Abhi Aiyer, Founder

Founder/CTO of Mastra. Principal eng & lead of >100 person eng org at Netlify. Built Gatsby Cloud infra that ran 10s of thousands of build nodes, served billions of files, and 100s of TB of monthly bandwidth.
Abhi Aiyer
Abhi Aiyer
Mastra

Shane Thomas, Founder

Founder/CPO of Mastra. Prev: built Audiofeed (AI podcasting). Staff eng / head of product @ Gatsby. 15+ years in open source. State wrestling champion.
Shane Thomas
Shane Thomas
Mastra
Company Launches
Mastra: the open-source JavaScript framework for building agents
See original launch post ›

Hey everyone 🙂 we’re Sam, Shane and Abhi.

What we’re building: Mastra (https://mastra.ai) is an open-source JavaScript SDK for building agents on top of Vercel’s AI SDK.

https://www.youtube.com/watch?v=8o_Ejbcw5s8

Our backstory: Previously, we built Gatsby, the open-source React web framework.

We noticed our friends building AI applications getting stuck debugging prompts, figuring out why their agents called (or didn’t call) tools, and writing lots of custom memory retrieval logic.

In October, we started building Mastra, and last week we went viral on Hacker News and Github and went from 1,500 to 7,500 stars.

Mastra use-cases: Fireworks AI is using us to build agentic workflows. YC companies are using Mastra to automate support, build CAD diagrams, scrape the web for contact info, automate medical transcriptions, generate financial documents, and create code generation products.

Mastra allows you to create workflow graphs that can suspend/resume, build RAG pipelines, write evals, give agents memory, create multi-agent workflows, and test it all in a local playground.

Here’s a weather agent and the agent playground (with tracing built in!):

You can deploy Mastra agents and workflows to a serverless platform like Vercel, Cloudflare, Netlify, any cloud service that can run Node.js, or our own hosting platform with evals and tracing.

How to get started: Just head over to the command line and install:npm create mastra

Happy building 🙂

PS: we’ve spent a lot of time this batch doing AI engineering whiteboarding sessions with other batch startups, walking through architecture and best practices, so we decided to write a book to help people get up to speed faster: Principles of Building AI Agents.

Follow @mastra_ai on Twitter/X and tweet at us and we’ll get you a physical copy!