Applied AI Lead
About the role
Covecta's ambition is to help financial institutions navigate and capitalise on the next industrial revolution, to establish itself as both the trusted partner of choice and the builder of an category-defining Agentic OS that banks depend on, extend and trust.
AI capability is developing fast, but capability that is not adopted by the teams it could benefit delivers nothing. The value only lands when AI is deployed inside real work and trusted by the people who do that work. 'Applied AI' is Covecta's customer arm and one of the most critical functions in the company. It owns the journey from signed contract to embedded, measurable production value. In regulated lending that gap is widest, because the blocker is as much a trust and change problem as an engineering one.
We are looking for an Applied AI Lead who blends banking domain credibility, hands-on agentic build skill, and the influencing and coaching ability to drive adoption. This role builds and proves the AI; it does not just promote it.
Who we are looking for
- Banking credibility. You can sit with credit officers and relationship managers and understand how the work really gets done.
- A builder, not an adviser. You configure the workflow yourself, then prove it’s right against real evidence before anyone relies on it.
- A coach. You can move a team from doing the work to reviewing the AI, and make the change stick.
- Fluent in AI. You already use AI tools well and can show where you’ve deployed them at work.
- Fast and accountable. You own the outcome and move at start-up pace.
What you will do
- Translate: Turn how the bank works into an agent specification, including the unwritten rules and edge cases.
- Build: Configure and tune agentic workflows on the Covecta platform, including prompts, judges, outputs and decision boundaries.
- Prove: Define acceptance criteria with bankers, build eval sets, and hold the quality bar before go-live.
- Drive adoption: Stand up champions and super users inside the bank, and grow usage that lasts.
- Show the value: Report the hours returned and capacity unlocked, surface expansion opportunities, and feed what you learn back into the product.
