now booking Q3 sprints

The AI Software Factory —
write specs, ship products_

You write a Markdown spec. We turn it into a production app — fast, opinionated, and built with AI agents + senior engineers who actually ship.

$ start_project --nowsee how it works →
spec.md
# Product: TicketDuck
> A SaaS that turns customer emails into tickets.

## Stack
- Next.js 15 + tRPC + Postgres
- Auth: Clerk
- AI: Claude for triage + summary

## MVP scope
- [x] Inbox sync (Gmail, IMAP)
- [x] Auto-tag with priority + topic
- [x] Slack notifications
- [ ] Billing (Stripe) — phase 2

## Done =
- Deployed on Vercel
- 90% test coverage on core
- Docs at /docs
$ ai-factory build → shipping…

// teams already shipping with us

◆ Acme
◆ Globex
◆ Initech
◆ Hooli
◆ Stark
◆ Wayne
// the problem

Software is built once, but explained a thousand times.

Slack threads, half-finished Notion pages, three different Figmas. Then six months later: nobody remembers why the auth flow exists.

// the old way
  • × 200-page PRDs nobody reads
  • × Devs guessing intent from screenshots
  • × "Final", "final-v2", "final-FINAL"
  • × Knowledge living in one engineer's head
// the .md way
  • ✓ One Markdown spec = source of truth
  • ✓ Versioned in git, reviewed in PRs
  • ✓ AI agents read it. Engineers ship it.
  • ✓ Docs are never out of date — they're the brief
// how it works

Four steps. No surprises.

Pick a sprint, hand us the spec, watch it ship. We optimise for momentum, not meetings.

01

Spec

We co-write a `spec.md` with you. Goals, constraints, success metrics. No 100-page docs.

$ ai-factory init
02

Plan

AI agents draft architecture, milestones, and risks. Senior engineers review every line.

$ ai-factory plan
03

Build

Daily commits, weekly demos, public Linear board. You see progress in real time.

$ ai-factory build --watch
04

Ship

Deployed on Vercel/Fly/AWS. CI, monitoring, docs included. We hand over the keys.

$ ai-factory ship --prod
// services

What we build.

If it can be specced in Markdown, we can probably ship it.

$ build --mvp

MVP Sprints

0 → launched in 4–8 weeks. Fixed scope, fixed price.

$ build --ai

AI Integrations

RAG, agents, tool-use. Plug AI into your existing product.

$ build --web

Web Apps

Full-stack Next.js, TypeScript, Postgres. Boring, reliable, fast.

$ audit --deep

Audit & Strategy

Codebase review, AI roadmap, hire/build advice. 1–2 week engagements.

$ team --extend

Staff Augmentation

Drop in a senior engineer (+ AI co-pilot) for 3–6 months.

$ build --internal

Internal Tools

Dashboards, ops tools, automations. The boring stuff that saves hours.

// stack

Boring tech, modern AI.

We don't chase frameworks. We ship on the stack that gets out of the way.

Next.jsTypeScriptPythonPostgrestRPCTailwindClaudeOpenAILangGraphVercelSupabaseStripeClerkDrizzlePrismaRedisCloudflareAWS
// selected work

Some things we built.

Names changed because NDAs. Outcomes are real.

MVP · 6 weeks

TicketDuck

AI inbox triage for support teams. Shipped to first 50 paying customers in week 7.

read case study →
AI · 4 weeks

LedgerLens

RAG over 10y of financial filings. Cut research time from hours to minutes.

read case study →
Internal · 3 weeks

OpsBoard

Replaced 6 spreadsheets with one dashboard. Saved a 2-person ops team 12h/week.

read case study →
// pricing

Pick a shape, not a contract.

Fixed scope, fixed price. No surprise invoices. Custom on request.

$ ai-factory plan --sprint

Sprint

from $8k

1–2 weeks. Audit, prototype, or focused feature.

  • ✓Discovery + spec.md
  • ✓Working prototype
  • ✓Loom walkthrough
  • ✓1 senior engineer
$ choose --sprint
$ ai-factory plan --build

Build

from $25k

4–8 weeks. MVP from zero to live.

  • ✓Full spec + roadmap
  • ✓Production deploy
  • ✓CI, tests, docs
  • ✓2-3 engineers + AI agents
  • ✓Weekly demos
$ choose --build
$ ai-factory plan --partner

Partner

monthly

3–12 months. Embedded team, shared roadmap.

  • ✓Dedicated squad
  • ✓Slack-first ops
  • ✓On-call & monitoring
  • ✓Quarterly strategy
  • ✓Custom SLAs
$ choose --partner
// what people say

Receipts.

“They shipped in 5 weeks what our previous agency couldn't in 6 months. The Markdown spec thing actually works.”

— Founder, Series A SaaS

“Felt like having a senior team on tap. Daily commits, no babysitting.”

— CTO, B2B fintech

“Best part: when we asked 'why', they pointed to the spec. Not a meeting.”

— PM, healthtech startup
// faq

Stuff people ask.

›Why .md as a domain?

Because Markdown specs are how we work. The TLD is a wink — and yes, it's a real ccTLD (Moldova). Works on every browser.

›Do you actually use AI, or is this a buzzword?

Real AI. Claude + custom agents inside our pipeline. They draft, refactor, and review. Humans own architecture and ship decisions.

›Who owns the code?

You do. 100%. We hand over the GitHub repo, infra access, and docs on day one of go-live.

›Can you start next week?

Usually yes for Sprints. Build engagements need a 1–2 week lead time so we can scope properly.

›Where are you based?

Distributed team across EU + US time zones. Most clients are async-first; we adapt to yours.

›What if I just need an audit?

That's the Sprint tier. We'll review your codebase, AI strategy, or product, and ship a written report + Loom in 1–2 weeks.

// next step

Got a spec? Got a napkin sketch?

Either works. Tell us what you want shipped, we'll tell you how fast and how much.

$ start_project --nowhello@aifactory.md