AI Advisory & Workshops for Engineering Teams

Rob Conery

Your company already tried AI. I make it actually pay off.

I built the AI and Copilot workshops for Microsoft's VS Code team, and before that I was the ASP.NET MVC Community Lead. Now I do two things: I work as a standing AI advisor to engineering leadership, and I run hands-on workshops inside the code teams actually ship. The hardest version of this problem lives in big enterprise .NET shops, which is where I spend a lot of my time, but nothing about the practices is locked to it. No hype. No slides about "the future." Just work that ships.

Rob Conery on the beach

Three ways in. Every price is right here on the page.

An ongoing advisor for the strategy, a two-day workshop for the team, or a half day to find out where you actually stand. Built for shops with real constraints: a decade of legacy code, a compliance team, and an agent that just opened a 4,000-file pull request.

Two engineers pair programming on real code
The Workshop

Agentic .NET: Bringing an Enterprise Team to 2026

Two days · on-site or remote · your solution

$12,500 fixed price, all-in

Two days of hands-on work with Claude Code inside the .NET code your team actually ships. Legacy modernization, agent-scale pull requests, review practices that hold up in a regulated shop, and the human part nobody plans for: getting a team of twenty to change how they work.

Your team walks out with:

  • A real modernization slice done together: Framework 4.x code moved onto .NET 10, tests and all
  • Claude Code wired into your solution, MSBuild, NuGet feeds, and CI, with house rules checked into the repo
  • A review playbook for the 4,000-file PR, so nobody rubber-stamps and nobody stalls
  • A 90-day rollout plan your leadership and your security team can both defend
Included: 10 team seats in The AI Pro, my private community for working engineers. The learning doesn't stop when I log off.
Book the workshop
The Front Door

.NET Modernization Assessment

Half day · remote · report included

$4,500 credits toward the workshop

Not sure where to start? Spend a half day with me on your actual solution and find out before you commit to anything bigger.

What you get:

  • A working review of your solution: framework versions, dead projects, test coverage, build pipeline
  • Which upgrade path is real for you and which one is a two-year sinkhole
  • Where agents will pay off first in your code, and where they'll bite you
  • A written report with a clear recommendation, sized for a leadership audience
The math: every dollar of the assessment credits toward the workshop. If we move forward, it cost you nothing.
Book an assessment

The price on this page is the price on the proposal. No "it depends," no discovery-call markup. Fixed, public, boring.

Six enterprise problems, two days, your code.

The examples are .NET, because that's where the hardest version of this problem lives: old code, big teams, and an audit trail. The practices travel to whatever else you ship, and they're the same ones I bring to advisory work.

01

Upgrading legacy .NET

Framework 4.x, WebForms, WCF, and the project nobody wants to open. We use agents to map the dependency graph, carve out a slice that can actually move, and get it onto modern .NET with tests that prove behavior didn't change. Strangler patterns, not big-bang rewrites.

02

Using Claude, properly

Claude Code inside a real solution: multi-project builds, internal NuGet feeds, dotnet test, and the conventions your team argues about. Writing CLAUDE.md so it encodes your architecture instead of generic advice, plus subagents, plan mode, and when to just write it yourself.

03

The 4,000-file pull request

Agents produce more diff than humans can read, and enterprise teams hit this in week two. How to scope work so the PR arrives reviewable, how to split what already landed too big, and how to decide fast when a giant one shows up anyway.

04

Code review at scale

What review is for once a machine wrote the first draft. Reading for intent and blast radius instead of style, using agents as a first-pass reviewer without letting them approve their own work, and keeping a trail that satisfies whoever audits you.

05

Moving the team

The part that actually decides whether this works. Your senior people are skeptical for good reasons and your juniors are pasting output they don't understand. We work through both, plus what to measure, what to mandate, and what to leave alone.

06

Guardrails your auditors accept

Where code and context are allowed to go, what stays on your network, how to handle secrets and customer data in prompts, and what to write down so security signs off once instead of every sprint. Real policy, not a slide.

From email to shipped in four steps.

01

Email me

Tell me about your team and what you ship. I answer everything myself, usually within a day.

02

We scope it

One 30-minute call to figure out which of these you actually need, or whether you need me at all. Sometimes you don't, and I'd rather say so.

03

The work

Two days hands-on in your code, or a standing monthly seat at the table where the decisions get made. Either way, in your codebase and on your problems.

04

The follow-through

Your team's seats in The AI Pro keep the questions flowing and the momentum going after I'm gone. No cliff, no "good luck out there."

I've been in the room where it happened.

I spent three years on the VS Code team at Microsoft, where I built the AI and Copilot workshops and curriculum that taught the team behind the tools your engineers already use. Three more years helping enterprise teams get serious about PostgreSQL. My first run at Microsoft was as the ASP.NET MVC Community Lead, which means I've been arguing about .NET architecture with enterprise teams since before half your solution was written. Along the way I founded Tekpub, a developer screencast company that Pluralsight thought was worth acquiring.

I wrote The Imposter's Handbook because I'm largely self-taught and felt the weight of not knowing things I felt I should. Thirty thousand other developers apparently felt the same way.

Now I do one thing: I help engineering teams build AI as a capability, not evaluate it as a product. There's a real difference, and I've seen both sides of it. Sometimes that's two days in a room with your developers. More often now it's a standing seat at the table where the decisions get made, month after month.

Rob Conery surfing
10 Years at Microsoft
VS Code Built their AI workshops
30k+ Books sold
F500 Clients trained

CS fundamentals, no hand-waving.

30,000+ copies sold. Written for curious developers who want to understand the fundamentals without the academic gatekeeping.

Thinking out loud about code.

Fifteen years of posts on PostgreSQL, Elixir, Node.js, career, writing, and whatever I happened to find interesting that week.

Loading articles…

This Developer's Life

Co-hosted with Scott Hanselman. Stories about the human side of software, from getting fired to competitive drive to why we aren't on Mars yet.

I take a handful of clients at a time. That's it.

If you're a VP of Engineering or CTO trying to figure out what your company should actually do with AI (not the hype, the reality) let's talk. Whether it's a retainer, a workshop, an assessment, or you're just not sure yet, email me. I read and answer everything myself.