Oisin partners with leaders who want to move past the AI hype — and turn it into products that ship, scale, and earn customer trust.
What we do
Strategy that survives the first roadmap review. Integrations that hold up under real customer load. A product story your team can actually deliver.
From "we should use AI" to a sequenced plan with owners, budgets, and the honest constraints — regulatory, data, organizational — that will shape what you can actually ship.
Take a fuzzy AI idea from concept to a pilot customers will pay for. Human-in-the-loop workflows, LLM selection, evaluation harnesses, and the unglamorous plumbing that makes them reliable.
Wiring AI products into the CRM, ERP, payments, and identity systems your customers already run on — so AI is a feature of their operations, not a separate science project.
Selected work
A few of the engagements that shaped how Oisin thinks about AI, scale, and the messy middle between them.
Built customer success and product strategy for an AI-driven platform serving 1M+ learners across government and employer partners.
Read case study →Program-led the Wells Fargo–Wachovia consumer-web integration carrying combined traffic of 100M+ visits per month.
Read case study →Sold and led a Deloitte program to lift the in-house IT capability of a UK welfare institution — protecting $2.8B in outsourcing benefits across 120,000 staff.
Read case study →The interesting question is rarely "can AI do this?" It's whether your team, your data, and your customers are ready for the answer when it does. Oisin lives in that gap — between the demo and the operating model.
Approach
Oisin is the work of Jarlath O'Carroll, who has spent two decades inside the companies most consultants only describe — running engineering teams, owning P&L for product lines, negotiating with regulators, and integrating acquired companies the morning after the deal closed.
Engagements are deliberately small and senior. You work directly with someone who has carried a quota, missed a launch, hired and fired, and shipped AI to customers who care about uptime and accuracy more than novelty.
A 30-minute call is usually enough to know if Oisin is the right partner — or to point you to someone who is.
