AI agents are the new tool. Your problem is the same.
I embed in your team, find where agents fit, and own the result. You approve every important step.
25 years. Amazon, HomeStars, AutoTrader, Biiibo.
Track Record
25 years of this work. Where I've done it before.
Owned the workforce tools deployed across all global warehouses.
CTO through acquisition. Rebuilt the technology platform.
Eliminated daily production outages and tightened delivery.
Contract VP of Engineering for AutoSync. Led adoption of AI-assisted engineering workflows.
Sound Familiar?
These are the patterns I see in every scaling engineering organization.
Everything Takes Too Long
Simple features take months. Your competitors ship faster with smaller teams. Tech debt is slowing delivery.
A Tool Nobody Uses
You bought the licences. Your team is still "exploring." Nothing about how the work gets done has changed.
A Team Busy With Manual Work
Good engineers spend the week on chores. Reports, triage, and status updates take the time meant for real work.
How It Works
Four stages, in order. I measure first. Then I use AI to speed up the fixes at each stage. AI speeds up a healthy system. It also speeds up a broken one.
See
Measure first. Know where the team is and what blocks it.
Stabilize
Stop the fires. Fix what blocks delivery first.
Scale
Grow the team without breaking it. Better review, test, and release.
Accelerate
Use the newest tools where they help most, in the stable parts of the system.
The First 30 Days
- 1
Free Lite assessment. A short questionnaire, a one-page readout, and a 30-minute call. This is how every engagement starts. The AI Agent Audit is a separate free offer.
- 2
Days 1 to 10. I sit with your team, interview key people, and pull the data. I measure where you are and pick the two problems that hold you back most. By Day 14 the first fix ships.
- 3
Day 30. We compare the numbers with your baseline. You decide to continue or stop.
Three Results We Measure
We measure from your own baseline.
Business value
More of the roadmap ships. Revenue features reach customers sooner. Cloud cost goes down.
Measured by: % of roadmap delivered, cycle time, cost per customer
Team stability
Clear roles and owners. Less firefighting. Fewer people leave.
Measured by: KTLO %, regretted attrition, survey scores
Product stability
Fewer incidents and bugs. Faster recovery.
Measured by: Incidents per month, time to recover, bugs found by customers
For agent work we also measure hours saved each week, and AI spend per team.
What an agent can do in your company
AI agent integration works best one job at a time, with a person approving the steps that matter.
Engineering
Code review support
- Problem:
- Pull requests wait for review.
- Agent:
- Checks pull requests for common problems.
- Approval:
- A person decides what to merge.
Test writing
- Problem:
- Tests get skipped when time is short.
- Agent:
- Drafts tests.
- Approval:
- A person reviews them.
Incident triage
- Problem:
- In an incident, people dig through logs first.
- Agent:
- Gathers logs and drafts a summary.
- Approval:
- A person leads the response.
Release notes
- Problem:
- Release notes are written late, or not at all.
- Agent:
- Drafts notes from the changes.
- Approval:
- A person approves them before they go out.
Sales and CRM
CRM cleanup
- Problem:
- Duplicates and missing fields pile up.
- Agent:
- Finds them and proposes fixes.
- Approval:
- A person approves each batch.
Lead research
- Problem:
- Building a profile of each lead takes time.
- Agent:
- Builds the profile of each lead.
- Approval:
- A person picks who to contact.
Sales follow-up
- Problem:
- Follow-up emails wait because nobody has time to write them.
- Agent:
- Drafts them from CRM notes.
- Approval:
- A person approves each email.
Meeting notes to CRM
- Problem:
- Call notes never reach the CRM.
- Agent:
- Writes the call notes into the CRM.
- Approval:
- A person checks the summary.
Pipeline report
- Problem:
- The weekly pipeline summary takes manual work.
- Agent:
- Writes the weekly summary.
- Approval:
- A person reads it before it goes out.
Customer support and onboarding
Support triage
- Problem:
- Tickets pile up and common questions repeat.
- Agent:
- Sorts tickets and drafts replies for common questions.
- Approval:
- A person approves replies at first.
Customer onboarding
- Problem:
- New customers stall on missing documents.
- Agent:
- Collects documents and sends reminders.
- Approval:
- A person approves each reminder.
Operations
Weekly reports
- Problem:
- Someone pulls the same numbers every week.
- Agent:
- Pulls the numbers and writes a summary.
- Approval:
- A person reads it before it goes out.
Invoice follow-up
- Problem:
- Late invoices need reminders nobody sends.
- Agent:
- Drafts payment reminders.
- Approval:
- A person approves them.
Inbox sorting
- Problem:
- The inbox buries the messages that matter.
- Agent:
- Sorts email and drafts replies.
- Approval:
- A person approves replies before they send.
Heard about Grok Bot, OpenClaw, Hermes Agent, Meta Muse, or OpenAI Dots?
Good. The tools change every month. The question stays the same: does it fit your team, and what should it do first? I help you decide. I stay tool-neutral. I do not resell any tool.
OpenAI Dots • Meta Muse • Grok Bot • Hermes Agent • OpenClaw
Tool list as of October 2026.
Four questions I answer
- Is it safe for our data?
- Which job should it do first?
- What does it cost to run?
- When should we not use it?
Safety
- A person approves anything that leaves the company.
- Each agent gets its own account and the least access it needs.
- We check where each vendor stores data.
Price
I don't bill hourly. We agree on scope and targets upfront, so you know the cost before we start. Both run as a 90-day engagement.
- One day a week
- $10,000 CAD/month
- Two days a week
- $17,000 CAD/month
After Day 90 the work can continue at the same rate, or as an ongoing one-day-a-week operating partner role at $10,000 CAD/month.
Interim leadership, due diligence and fixed-scope projects have their own prices on the how I work page. All prices are in Canadian dollars (CAD).
The only guarantee: the Day-30 exit. You can stop after Day 30 and pay for Month 1 only. No conditions.
Frequently Asked Questions
Which AI tool do you use?
The best tool for the job. I stay tool-neutral and I do not resell any tool.
Do agents replace my engineers?
No. Agents speed up a healthy team. They also speed up a broken one, so we measure first.
Is it safe to give an agent access to our systems?
We limit access, and a person approves outside actions.
What does the free audit include?
Eight short questions about your team, mostly multiple choice. You get a one-page readout of five scores: delivery, manual work, data, AI use and ownership. We go through it on a 30-minute video call. The scores are self-reported, so they show where to look, not proof. It is free.
Is the AI Agent Audit the same as the Lite assessment?
No. The AI Agent Audit is a free, self-serve check of where AI agents fit your team. The Lite assessment is how every engagement starts: a short questionnaire, a one-page readout, and a 30-minute call. They are different offers.
How long does it take?
A 90-day engagement, one or two days a week. It starts with a free Lite assessment. In the first 10 days I measure where the team is. By Day 14 the first fix ships and we agree on targets. At Day 30 we check the numbers against the baseline, and you decide to continue or stop. From Day 31 I run the work at full depth and hand it to your own owners by Day 90. The only guarantee: the Day-30 exit. You can stop after Day 30 and pay for Month 1 only. No conditions.
What size companies do you work with?
I work with growth-stage companies, typically $5M–$50M in revenue with 10–200 people. If your team is busy but not shipping, tech debt is slowing delivery, or your AI tools never reached the workflow, that's exactly when I step in.