Christopher Brellis. Expertise
Contact

Expertise: B

Innovation & GenAI.

Turning GenAI from experiment into enterprise value: human-centred, governed, and trusted by customers, employees and regulators alike.

← What I do now

01 In short

Why do so many GenAI pilots never reach a customer?

GenAI pilots are easy. Ones that reach production in a regulated business are not.

As an EY AI Ambassador I work on GenAI-enabled transformation for European financial services. I design the human side (journeys, roles, controls and trust) so the technology lands as a service people rely on, not a demo.

02 How it works

Problem, approach, outcome.

Problem

Pilots that never ship.

Dozens of experiments. Few with an owner, a measurable job, or a path past risk and compliance.

Approach

Design the service, not the model.

  • Use cases ranked by value and risk
  • Conversation, agent and control design
  • The operating model: roles, skills, ways of working
Outcome

AI that customers and regulators trust.

Use cases in production, with clear accountability, measurable value and controls your risk team signs off.

03 What you get

  • Use-case portfolio — start where the value is proven
  • Chatbot & agent design — conversations that resolve, not deflect
  • AI governance & controls — sign-off from risk, early
  • GenAI operating model — a team ready for the next use case

04 Case studies

Selected work.

Financial services

A chatbot that passed compliance, and went live

Challenge
A financial-services client wanted AI customer support without regulatory or reputational risk.
What we did
Designed and deployed it: intents, conversations, hand-over to people, and guardrails agreed with risk.
Result
A live, governed service, not a pilot.

EY · internal

How agentic AI became an everyday working method

Challenge
Agentic AI was much discussed and little used in day-to-day work.
What we did
Created and championed agentic workflows in EY’s AI Ambassador community: demos, playbooks, coaching.
Result
Agentic AI taken up as a working method by teams.

Financial services · Europe

What GenAI changes inside a financial institution

Challenge
A European financial-services organisation needed to know how GenAI changes its roles, processes and controls.
What we did
Mapped where AI augments or automates, and designed governance and new ways of working.
Result
A target operating model and roadmap for adopting GenAI at scale.

Client names are withheld. Engagements delivered with EY Studio+ teams.

05 FAQ

Questions, answered.

Why do GenAI pilots stall?

Rarely because of the model. They stall on trust, workflow fit and accountability: no owner, no measure, no sign-off. Design makes those explicit before anything is built.

What is agentic AI?

AI that takes a goal, plans the steps and uses tools to complete them, with people supervising. Designing it means designing delegation: what the agent may do, when it must ask, and how its work is checked.

How do you handle risk and compliance?

From day one. Controls, human checkpoints and audit trails are designed with risk teams, not bolted on at the end.

Got a pilot that won’t ship?

Email hello@antidot.gr or message me on LinkedIn. I will bring the right EY team.

07 Contact

Let’s talk about your system.

Client work is delivered through EY Studio+. Write to me and I’ll bring the right team.

hello@antidot.gr

LinkedIn · antidot.gr