I help founder-led teams find the right AI opportunities, train their people, ship a measurable pilot, and keep improving what works.
built & advised
Worked inside these companies -- as a partner, product lead, and founder. Not as an outside vendor.
Every company I talk to wants AI agents. That's not the problem.
The real problem is simpler.
But those are symptoms.
The real blocker is the data. Not that it doesn't exist -- there's plenty. But 80% of what your company actually knows lives in Slack threads, email chains, WhatsApp messages, and calls nobody recorded. Unstructured. Unowned.
An AI agent doesn't fix that. It's just the first thing that runs into it.
That's why I start with a workshop or an audit, then choose one measurable pilot.
A practical workshop and setup session built around the team's real work.
Paid discovery that shows where AI can create a measurable business result.
One contained system tied to a KPI and used by the team within two weeks.
Ongoing ownership of your AI roadmap, implementation, adoption, and results.
We choose one workflow tied to growth, customer value, or cost. Then I build it with your team, measure the baseline, and leave a working system.
We walk through the workflow from start to finish, find the real constraint, choose one KPI, record the baseline, and define the result we want after 60 days.
I build the smallest useful system, test it on real work, train the person who will own it, and document what to improve next.
I work with 5-50 person teams that want AI to actually save time and money -- without a transformation project, an 80-page strategy document, or five discovery meetings before we start.
A short outside look at your processes and a clear answer: where AI will give you the most leverage for the least effort. No vendor sales pitch -- just a ranked opportunity list and a recommended first pilot you can ship in two weeks.
Four phases: diagnose, design, guide, review. We work on one workflow at a time, ship it to production in 2-4 weeks, then hand it off to an owner on your team. No consulting dependency once we're done -- your team runs it.
A five-dimension readiness score across process, data, team, budget, and culture. Twenty-five points total and a specific fix plan for your weakest dimension. One honest hour of scoring saves you months of 'why isn't this working?'
Tools I actually use every day. Copy, paste, run them in your AI.
Most teams do not have a ChatGPT problem. They have a context problem. AI needs your company rules, examples, process, and owner before it becomes useful every week.
GPT-5.5 and Claude Opus 4.7 both dropped in April 2026. I went through the benchmarks, developer tests, and community data to figure out which one your team should use.
Most AI outbound tools are expensive noise. Here is what practitioners are actually shipping: Claude as the outbound engine, MCP integrations that collapse five tools into one, and the signal-based approach that gets 5-7x higher reply rates than cold list email.
Live Artifacts turn Claude Cowork into a real-time dashboard layer. Pull Gmail, Stripe, Calendar, and Apify into one auto-refreshing view. Here are five I built for myself.
An AI partner works across the full path: team setup, opportunity audit, implementation, adoption, and measurement. We can start with one workshop or one focused pilot. If the work creates a measurable result, the relationship can continue as an ongoing fractional AI partnership.
Four phases: diagnose (map your workflows), design (pick the tool and write the prompts), guide (run it alongside the old process until it holds up), review (measure the outcome and pick the next workflow). One workflow per 90 days is a realistic pace for a small team. Two workflows in 90 days is excellent.
Interviews with 3-5 key people, a map of current processes, a list of 10-15 AI opportunities scored by hours-saved-per-week and setup effort, a recommended first pilot, and a 90-day action plan. All in two weeks. It's a working document for your team, not a consulting deck.
Each rung is scoped separately. A team setup is the smallest commitment. Audits and pilots use a fixed scope. An ongoing retainer starts after we have shipped useful work together. I give you a clear range and scope after the first call.
DIY if: 1-5 people, one painful workflow, one motivated person on the team, and a weekend to experiment. Hire an advisor if: 10+ people, AI has become 'everyone's job' and nothing sticks, three prior attempts bounced off, or you need an outside voice to tell the team what to stop doing. A good $5K advisor usually pays for themselves in the first month.
One production-ready workflow: 2-4 weeks from decision to live. Two workflows in 90 days: realistic first-quarter plan. A full team-wide shift to AI-assisted workflows: 6-12 months, but that's a culture change, not a project.
AI consulting is the outside advisor who decides what to do and in what order. AI implementation is the hands-on work of configuring the tool, writing the prompts, and getting it into production. A good engagement includes both: the consultant decides what, then personally implements one pilot by the end of two weeks and hands the rest to the team.
Yes, and it's usually the fastest win for a small business. AI marketing consulting works on your content process, email campaigns, lead research, competitive analysis, and first-draft ad copy. In two weeks you can stand up one or two automations that save a marketing team 5-10 hours a week.
If you're under 5 people with one painful workflow, no -- go straight to a pilot. If you're 10+ people, a few prior AI attempts have failed, and you can't decide what to tackle first, then yes. A good AI strategy isn't an 80-page deck -- it's a two-page list of what to do, in what order, and at what budget. Everything else is slides.
Yes. I work with founders and teams in the US, EU, and Ukraine. Everything runs remotely over video and async docs. I work in both English and Ukrainian. Time zones under a 6-hour difference are easiest to schedule.
It's normal and it's not your fault. Resistance disappears the moment one person shows real time savings on a real task. That's why the first pilot is always a boring, repetitive workflow -- not something ambitious. Once the team sees that the marketing lead now writes the weekly report in 15 minutes instead of 2 hours, resistance ends within a week.