One AI feature. One week. $2,500 flat.

Claude-powered chat, RAG, agent workflow, or MCP integration — built into your existing product. Working code in your repo, tests included, live demo on day five.

Reserve the week — $500 deposit Questions first? hello@novellogic.ca

One sprint slot per week.

Prices in USD.

What ships

Working feature merged into your repo — not a prototype, not a Loom
Test suite included, run before handoff
Claude API integration wired to your stack (auth, streaming, error paths)
Short written handoff doc: architecture, env vars, how to extend
Live demo walkthrough on day 5

The week

Day 1

Recon

Read your codebase, agree exact scope in writing.

Day 2–3

Build

Feature + tests, daily progress note.

Day 4

Integration

Your environment, your data, edge cases.

Day 5

Demo + handoff

Live walkthrough, docs, keys handed over.

Proof

Built on modules already running in production

90% judge pass · 30 cases
$0.004/case
41/41 tests
recall@3 1.00

See the modules →

Scope

In scope

One feature (chat, RAG, agent workflow, or MCP server/client integration)
Python or TypeScript
Your existing repo + infra
Anthropic API

Out of scope

Greenfield apps
UI redesigns
Multi-feature builds
Non-Claude LLM stacks
Ongoing retainers (see Advisory for that)

FAQ

What happens after I pay the deposit?

You get a scheduling email within one business day. We pick your week, you share repo access, day 1 starts with a scoping call. The $500 is applied to the $2,500 total; balance due at handoff.

What if the scope doesn't fit in a week?

Day 1 exists to prevent that. If we agree scope in writing and I miss it, the demo ships anyway and we finish it — the price doesn't change.

What do you need from me?

Repo access, an API key budget (~$20–50 of Anthropic usage for testing), and one 30-min call on day 1 and day 5. That's it.