01
Call one — build
90 minutes, screens shared. We pick the workflow that costs you the most and build the automation live, in your tools, with your data. You keep the file, the prompt, the credentials.
Product 01 / Advisev.2026.AI
The AI
Concierge.
Two calls a month where we build your AI workflows, automations and systems live with you — not for you. Direct access between calls. A deliberately small client list, so the calls stay senior.
The promise
Why a concierge
AI adoption is a culture problem, not a technology problem. Culture doesn't change in a one-off workshop — it changes when someone senior keeps showing up, building in the open, and making the new way of working normal.
The Concierge is that cadence, at its smallest useful size. Twice a month we sit down and build something real together. In between, you have a direct line. It's the cheapest way to keep momentum without hiring a team you don't need yet.
How a month runs
Cadence
02
Between calls — direct access
Direct messaging with us for the questions that would otherwise stall a week: which tool, which model, is this prompt safe, is this output trustworthy, why did this break.
03
Call two — extend & hand over
We harden what we built, hand it to the person who will own it, and pick the next one. Every month adds another working system, not another slide.
What's included
Every month
- 01
Two live build sessions per month (workflows, automations, agents, internal tools).
- 02
Direct messaging access between calls — no ticket queue.
- 03
A running backlog of your automation opportunities, ranked by hours saved.
- 04
Every artefact documented and owned by you, on your accounts.
- 05
Responsible-AI module built in: data-handling rules, approved-tool list, human-in-the-loop checks and a shadow-AI watchlist reviewed each month.
- 06
A quarterly written note on what changed, what it saved, and what to do next.
Why the client list is capped
Scarcity, honestly
We hold a small number of Concierge clients at any time — because the person on your call is the person who has actually built and scaled this inside a company, not a junior with a deck.
No pooled account managers, no rotating bench, no reselling a vendor's licences. If the slots are full we'll tell you, and offer a Sprint instead.
Outcomes
What changes
2
Live builds / month
1
Named internal owner
<24h
Answer between calls
6
Clients, capped
- Leaders stop asking whether AI applies to them and start bringing workflows to the call.
- The team's first instinct on a repetitive task becomes "can we automate this before Friday?"
- Tool decisions get made in a day instead of a quarter, against your criteria rather than a vendor's.
- Shadow AI turns into sanctioned AI, because there is a monthly forum where it surfaces safely.
The team who runs it
The bench
Led by an operator. Staffed by practitioners who've shipped AI in production.
Every engagement is led by Joao Alves, a second-time founder who has worked in AI since 2020 — most recently inside a leading fintech company at a $7B+ valuation doing over $500M ARR, helping regulated firms put AI into production.
Joao is supported by a small bench of senior practitioners drawn from financial services, VC-backed AI startups, and risk and governance teams. Engagements are staffed for fit — every workshop, team challenge and governance program is led by people who have done the work, not delegated it.
- AI operators since 2020
- Fintech, regulated industries
- Shipped, not just advised
- Workshops on 3 continents
- Bench scales with the brief
Next step
When one team's workflow isn't the whole problem, the Sprint goes wider.
Two calls a month is enough to start
If the slots are open, we'll say so on the first call. If they aren't, we'll tell you that too.