Purpose-built
Every system is shaped around the work it has to do — not a template, not a demo, and not a wrapper around someone else's product.
NewSwarm extraction in openextract-ts
Production AI systems for teams and agents
Purpose-built for the difficult parts — extraction, agents, evaluation, and the infrastructure underneath. Built in the open. Owned by you.
The stack we build on
- TypeScript
- Python
- Rust
- Next.js
- AI SDK
- pydantic-ai
- Postgres
- MCP
- Zod
- Vercel
A different kind of AI partner. Small senior teams, shipping systems you keep — open wherever it can be.
Mellow AI is an independent AI engineering studio in Dallas, Texas, working remotely with teams worldwide. We design and ship production AI systems — document extraction, agents, evaluation, and the infrastructure underneath — inside your environment, and hand over every piece: the code, the models, the infrastructure, and the operating knowledge.
Powered by models, not guesses
Multiple models read the same input and reconcile their answers. Confidence, provenance, and the source clause travel with every field.
Owned by you
The code, the models, the infrastructure, and the operating knowledge are yours from day one. No lock-in, no black boxes.
Selected systems
Useful AI, built all the way through
Four systems we build and maintain in the open. Nothing here is a mockup — read the source, run the benchmark, install the package yourself.
Open source / TypeScript01
openextract-ts
The TypeScript line of openextract. One call takes a file to a validated Zod object — and a swarm runs several models over the same document, then reconciles their answers by merge, vote, or first success. An MCP server, a CLI, a terminal UI, and a local web cookbook ship with it.
- Runtime
- Node 20+, TypeScript
- Swarm reduce
- merge, vote, or first
- Also ships
- MCP server, CLI, TUI, web UI
- Test coverage
- 100%, enforced in CI
- License
- MIT
Open source / Python02
openextract
Documents, images, audio, and video into a validated Pydantic model in one function call. Bring your own model — eleven providers work out of the box, and the same schema holds across all of them.
- Runtime
- Python 3.12+
- Providers
- 11, via pydantic-ai
- Test coverage
- 100%, enforced in CI
- License
- MIT
Applied system / TypeScript03
Open Lease Audit
A portfolio integrity console for commercial leases. Define the fields that matter, stream an abstraction of every document, then reconcile the whole portfolio for conflicting terms and contradictions.
- Per field
- Source clause + confidence
- Findings
- Ranked by severity
- Intake
- PDF, text, Markdown
- License
- MIT
Open source / Rust04
pdfmd
PDF to Markdown with nothing underneath it. The object-graph reader, the DEFLATE decoder, the font decoding, and the column and table heuristics all live in the crate.
- Dependencies
- 0
- Throughput
- ~3,900 pages/sec
- Test coverage
- 99%
- License
- MIT
How we work
Your systems. Your edge.
We build inside your environment, not a demo. Every decision is made to create durable capability rather than dependency.
Start a conversation- 01
Find the leverage
We map the system, pressure-test the opportunity, and define the smallest valuable release.
- 02
Build the real thing
Senior engineers work directly with your team and ship into your environment — not a sandbox.
- 03
Leave you stronger
The code, models, infrastructure, and operating knowledge are yours from day one.
Questions
What people ask before they call
Short answers, written to be read by a person or quoted by an assistant.
What is Mellow AI?
Mellow AI is an independent AI engineering studio based in Dallas, Texas, working remotely with clients worldwide. Mellow AI designs and ships production AI systems — document extraction, agents, evaluation, and the infrastructure underneath — for teams with ambitious, difficult problems.
Who owns the code Mellow AI builds?
The client does, from day one. The code, the models, the infrastructure, and the operating knowledge belong to the client. Mellow AI does not sell a platform, resell a license, or hold anything back — there is no lock-in and no black box.
What kinds of projects does Mellow AI take on?
Typical engagements are custom AI systems, agent and tool-use architectures, LLM deployment, document and data extraction, evaluation harnesses, model fine-tuning, AI project rescue, and the production infrastructure underneath them. Mellow AI builds inside the client's environment rather than in a sandbox or a demo.
What open-source software does Mellow AI maintain?
Four public systems, all MIT-licensed: openextract-ts, a TypeScript extraction toolkit with multi-model swarm reconciliation; openextract, the Python line, covering 11 model providers through pydantic-ai; Open Lease Audit, a portfolio integrity console for commercial leases; and pdfmd, a zero-dependency Rust PDF-to-Markdown converter that averages 4.4 ms on a 17-page paper.
What is swarm extraction?
Swarm extraction runs several models over the same document and reconciles their answers field by field — by merge, by majority vote, or by first success. Every field carries its confidence grade and the source clause it came from, so a low-confidence value can be routed to human review instead of silently entering a ledger.
What technologies does Mellow AI build with?
TypeScript, Python, and Rust, on Next.js, the Vercel AI SDK, pydantic-ai, Postgres, Zod, and the Model Context Protocol (MCP), deployed on Vercel. Model choice is deliberately not fixed: Mellow AI builds systems that run across providers rather than against a single vendor's API.
How does an engagement with Mellow AI start?
It starts with a conversation about where the work is slow, brittle, or impossible today. Mellow AI then maps the system, defines the smallest valuable release, and ships it with senior engineers working directly alongside the client's team. Email info@staymellow.ai or use the form at staymellow.ai/contact — replies come within one business day.
Can an AI assistant or agent query Mellow AI directly?
Yes. Mellow AI runs a public Model Context Protocol (MCP) server at staymellow.ai/api/mcp over streamable HTTP. Any MCP client — Claude Code, Cursor, or another agent — can read the studio's facts and call its model-aware prompt optimization tool. A plain-text summary for language models is published at staymellow.ai/llms.txt.
Where is Mellow AI located?
Mellow AI is based in Dallas, Texas, in the United States, and works remotely with clients worldwide.
Building with an assistant? Connect our MCP server or read the plain-text brief at /llms.txt.
Have a hard problem?
Let's build what should exist
Tell us where the work is slow, brittle, or impossible today. We'll respond with a clear point of view.
The Mellow Signal
A useful signal in the noise
Brief, opinionated notes on models, agents, infrastructure, and the work of putting AI into production.
No noise. Unsubscribe anytime.