- Developed an LLM-powered entity research service using the Gemini API for automated adverse-media and risk profiling; integrated into production compliance workflows.
- Designed and built an automated enterprise-wide risk assessment tool that processes compliance data and generates structured Word and Excel reports, reducing manual reporting effort.
- Built a Kubernetes-native Root Cause Analysis service that ingests Grafana alerts and autonomously investigates incidents using LLM agents; extended it with a purpose-built Kanban board tool for AI-native development workflows.
- Led the deployment of a sanctions list management service across dev, staging, and production environments.
- Coached 4+ internal hackathons (2-day and 2-week formats) on AI-assisted development; co-developed a full compliance platform frontend (TypeScript/Angular) alongside the coached team.
- Coached a UI test automation hackathon: set up the end-to-end (E2E) testing infrastructure — Playwright test suites running in GitHub Actions against a sanctions screening platform's UI, with a custom dashboard visualising test results — now adopted and in active use by the team.
- Presented AI engineering work and platform capabilities to internal audiences of ~500 across multiple online forums and company-wide events.
- Delivered hands-on AI tooling workshops for colleagues covering IDE-integrated AI assistants, Jira integration via the Model Context Protocol (MCP), and agent/skill definitions.
- Active contributor to the company-wide initiative to redesign the SDLC (Software Development Lifecycle) around AI-native practices, including architectural design input for a new internal AI coding platform.