Engineering firms
In-house calculators, report generators, and analysis tools that your engineers use daily but nobody maintains.
- Civil
- Mechanical
- Electrical
- Structural
An AI Consultancy
AI for the codebase
you
already have.
We audit the small codebases that non-software teams quietly depend on, then retrofit them so AI can do real work inside them. We stay until your team runs the loop without us.
- Engagement
- 6–8 weeks
- Team size
- 1–10 engineers
- Deliverable
- Checked-in config
- After handoff
- Retained
Who we work with
Small teams that aren't software companies but depend on software anyway.
You inherited a codebase. It isn't your business, but your business runs on it now. The person who wrote it moved on, the AI tooling changes every month, and you don't have a quarter of meetings to figure out what to do about either.
Specialty consultancies
Custom models, pricing tools, and client deliverables that started as spreadsheets and outgrew them.
- Actuarial
- Environmental
- Biotech
- Policy
Research & in-house teams
Data pipelines, experiment runners, and one-off scripts that quietly became infrastructure.
- Labs
- Nonprofits
- Operations
- Finance
Services
What we deliver.
Three engagements, roughly in the order teams buy them. Most start with the diagnostic and end up on the retainer.
2 weeks
Codebase audit
We read every line that matters. You get a written report with an AI-readiness score, a critical-path map, and a prioritized list of what to do next.
- Dependency + dead-code pass
- Structural documentation you actually have
- AI-readiness score across eight dimensions
- Prioritized remediation plan
4–6 weeks
AI enablement
We retrofit the repo for agents: structured memory, prompts that actually work, role-specific workflows. Then we sit with each seat until it sticks.
- Codebase memory + rules configured
- Claude Code / Cursor / agent config checked in
- Playbooks for PMs, engineers, and reviewers
- Hands-on workshops, role by role
Ongoing
Retained advisory
After handoff, we stay reachable at a reduced rate. Office hours, PR review on agent-generated changes, and triage when a new model changes the job.
- Weekly office hours
- Review queue for AI-generated PRs
- Model-change triage & migration
- Month-to-month, cancel anytime
Approach
Diagnose. Build. Partner.
We don't hand you a slide deck and leave. The engagement is shaped so your team owns the loop by the end.
- Week 1–2
Diagnose
We read the code, interview the team, and map how work actually moves through the repo. You get a written report that says what's there, what's missing, and what AI can do with the codebase today.
- Week 3–8
Build & teach
We retrofit the repo for agents, check in the config and prompts, and sit with each seat until they can drive the new loop without us. No black boxes. Everything we add stays checked into your repo.
- Month 3+
Partnership
Month-to-month retainer at a reduced rate. Weekly office hours, PR review on agent-authored changes, and someone on call when a model update changes the job.
Pricing
Transparent, fixed-scope.
No hourly billing, no surprise invoices. Scope and price are agreed up front, in writing. The retainer is month-to-month, so you can leave when you don't need us.
Diagnostic
Two-week read of the codebase and a written report with an AI-readiness score and a prioritized plan. Standalone or credited toward Engagement.
$8k Fixed
- Full codebase read
- AI-readiness scorecard
- Written report (20–40 pages)
- 1-hour walkthrough call
Most common
Engagement
End-to-end AI enablement of the repo plus training for your team. Price depends on repo size, team size, and scope agreed up front.
$35–60k Fixed
- Everything in Diagnostic
- Agent config + codebase memory
- Role-specific playbooks
- Hands-on training sessions
- First 30 days of Retained included
Retained
Ongoing advisory at a rate that works for teams who aren't software companies. Month-to-month. Cancel or pause anytime. For teams that finished Engagement, or anyone who wants a steady hand when the model changes.
$2.5k / month
- Weekly 30-min office hours
- PR review on agent-authored changes
- Model-change triage
- Priority email response
Prices are starting points. We publish them so you can qualify us out before the first call. Final quotes are written after a 30-minute intro.
Contact
Tell us about your codebase.
A paragraph is enough. What it does, how old it is, who maintains it, and what you wish it did. We'll reply within a day with either a meeting slot or a polite referral.
[email protected]
24-hour response, weekdays.