An engineer inside your team for work that doesn't fit a fixed package: an AI feature that has to fit your codebase, requirements that move week to week, or extra hands on your roadmap.

Software Engineer: $65–100/hour. Product and platform work in TypeScript, Python and Next.js.

AI Operator: $65–85/hour. Running, reviewing and tuning agents and automations, handling exceptions and reporting.

Forward-Deployed Engineer: $90–125/hour. Getting an AI system from pilot into production in your environment.

AI Systems Engineer: $90–130/hour. Agent architecture, evaluation, retrieval and infrastructure.

Technical Consultant: custom quote. Reviews, architecture decisions and model or vendor choices.

The rate sits lower in the range for reserved capacity and clear scope, and higher for ad-hoc hours and tight deadlines. We agree it before work starts.

Hourly: a first block of at least 10 hours, then as needed

Fixed blocks: 20 or 40 hours, prepaid, used within 60 days. Unused hours are refunded.

Monthly: Engineering Support from $1,000/month for 12 reserved hours

Agency subcontracting: I work under your agency's brand or alongside your team

Technical operations: the AI Operator track, or Managed Automation from $500/month

Availability: weekly hours and an overlap window with your core hours, written into the agreement

Time reporting: a weekly time log in 15-minute increments, by task

Deliverables: code, docs and tests in your repositories through your normal review process

IP: you own the work product I create for you once it's paid for. I keep my pre-existing generic tools and tell you before I use any.

Access: your accounts, least privilege, no shared passwords. Access is removed at the end.

Communication: a weekly written update and an optional weekly call

Response: messages answered within one working day. No on-call or out-of-hours cover unless priced.

Termination: 7 days' written notice for hourly and weekly work, 30 days for monthly. Hours worked are paid; unused prepaid hours are refunded.

Change requests: hourly work follows your priorities. Fixed blocks change only by written agreement.

On-call or out-of-hours cover unless priced, third-party tools, hosting and API fees, and the obligations of an employee. I work as an independent contractor.

An intro call about the work, your stack and how your team operates

One person who sets priorities and reviews my work

Access through accounts you control

A first goal for the first two weeks

Each week, the update shows what was agreed, what shipped and what's blocked, and the time log matches the invoice. At the end, you get the handover note.

You choose the length. The first block is at least 10 hours, reserved weekly capacity runs in 4-week minimums, and fixed blocks are used within 60 days. We agree the start date and weekly hours in writing after the intro call.

If the work grows past the first goal, we agree the next block or move to reserved weekly capacity. If it turns out to need forward-deployed or AI systems work rather than general engineering, I say so and we agree the track and rate in writing before those hours start. The rate never changes on hours already agreed.

Well-defined pieces are often cheaper as packages: AI Workflow Pilot $750, Repository Health Review $400, Premium Landing Page $750, Self-Hosted Agent Setup $600.

AI Agent Run Debugger (Prototype: new screens for my public Python runtime, which has 525 passing tests; runs shown are illustrative)

Codebase Audit of an Open-Source Python Runtime (Implemented: a real audit of my own public repository, every finding cited to a file and line)

AI Agent Mission Control (Concept: an interface for two orchestration specs I wrote; no code behind it yet; illustrative data)

Message me on Contra with the work, your stack and the hours a week you need. I'll reply within two working days to set up an intro call.