Associate Director of Applied AI/ML at Fitch Ratings. Building production AI systems, leading engineering teams, and bridging the gap between research and real-world deployment.
engineers led
8+
products shipped
15+
devs mentored
20+
industries
4
Analyzing the shift towards small language models, local inference, and the rise of sovereign AI infrastructure.
How our team built a winning DeFi application during the Ethena Hackathon, covering technical challenges and lessons learned.
The motivations, challenges, and skill transfers that led me from data science into full stack development and entrepreneurship.

current focus
Agentic AI, RAG, and LLMOps for high-trust environments
Architecture, implementation, and delivery from prototype to production.
About

I lead applied AI and machine learning at Fitch Ratings, where I design and build agentic AI systems, RAG pipelines, and production ML infrastructure that drive credit ratings analytics at global scale. My work sits at the intersection of research and engineering - translating frontier capabilities into reliable, regulated financial systems.
Beyond individual contribution, I lead engineering teams, define technical roadmaps, and mentor developers through complex cross-functional delivery. I believe that great technology is built by empowered teams with clear purpose, and I invest heavily in creating that environment. I also contribute to open source projects and maintain tools that help other engineers ship better AI systems.
MSc Data Science, University of Bristol - with a focus on machine learning, data analysis, and advanced algorithms.
Associate Director, Applied AI/ML
Agentic AI, RAG pipelines, LLMOps for credit ratings analytics
Full Stack Python Developer
Trade compression systems, Flask APIs, Angular interfaces
Python Developer
ML pipelines and pharma data transformation workflows
Software Development Intern
RPA monitoring tools, Spring Boot, Angular, Elasticsearch
capability matrix
The stack is organized around how I actually deliver work: applied AI, backend systems, data pipelines, and infrastructure that can be operated by a team.
I care about the whole path from prototype to production: model choice, orchestration, APIs, deployment, observability, and the developer ergonomics around it.
delivery style
production-first
primary focus
applied ai
supporting stack
full stack + infra
production output
A portfolio shaped around outcomes, architecture, and delivery rather than just job titles.
Fitch Ratings|2026 - present
End-to-end agentic AI systems and RAG pipelines enabling analysts to interact with credit data through natural language.
agentic ai
retrieval
production
BNP Paribas|2021 - 2026
Python automation and Flask APIs for multi-billion dollar swap compressions.
workflow automation
enterprise
fintech
Definitive Healthcare|2019 - 2020
Data transformation and predictive ML models for pharmaceutical analytics.
ml pipelines
analytics
data quality
Principal Financial Group|2019
Full stack RPA monitoring tools for large-scale enterprise DevOps.
monitoring
automation
internal tooling
Multi-agent workflows and autonomous research agents using LangChain and LangGraph for production financial analytics.
Document intelligence pipelines with embedding models and Azure AI Search for semantic retrieval over large-scale corpora.
End-to-end LLM lifecycle: prompt versioning, evaluation frameworks, experiment tracking, and production monitoring.
End-to-end extraction and analysis with pymupdf4llm and vector search for financial filings.
approach
Process, methodology & engineering leadership in practice
Structured Logging
Structured logs for every LLM call and pipeline step - full auditability in regulated environments
Latency Monitoring
Per-model and per-endpoint latency tracking with alerting for degradation
Cost Tracking
Token usage monitoring and budget alerting across AWS Bedrock and Azure AI Foundry
Answer Quality
Automated regression benchmarks before every deployment, A/B evaluation frameworks
Error Handling
Fallback strategies, retry logic, and graceful degradation for LLM call failures
Pipeline Orchestration
ML and data pipelines orchestrated with Airflow, SageMaker, and Celery across hybrid cloud infrastructure
Distributed Task Processing
Async ML inference and high-throughput API serving with Redis and Celery for production workloads
CI/CD for ML
Automated testing, model evaluation, and deployment pipelines with GitHub Actions, Docker, and Kubernetes
Experiment Tracking
Rigorous A/B evaluation frameworks and systematic quality benchmarks before every production rollout
Infrastructure Stack
Deploying across AWS (ECS, Lambda, S3, SageMaker), Azure AI Foundry, bare metal (Hetzner, Digital Ocean), and hybrid infrastructure.
Building production AI systems in financial services - the lessons, failures, and patterns that actually work at scale
Share applied AI engineering learnings at conferences, tech meetups, and Python/JS community events in London and beyond
Experiment with new frameworks, contribute to open source, and share what gets learned along the way
Grow from technical lead into engineering manager, building and leading a high-performing applied AI team
latest activity
work with me
recognition
Awards, mentoring, and the drive to share what works
Hackathon wins and professional credentials
Ethena Hackathon Winner
1st placeFirst place at Ethena Hackathon with the E-Sky project - a real-time on-chain token explorer built with Next.js and Goldsky.
ICP Hackathon Finalist
finalistReached the finals at Internet Computer Protocol Hackathon with nexBit - a Bitcoin explorer built with Rust and Internet Identity.
Professional Scrum Master
certifiedPSM certification for agile engineering practices and team delivery management.
Interest Rate Swaps (LFS)
certifiedSpecialised certification in financial derivatives and trading systems for fixed-income markets.
Engineering leadership, community, and knowledge sharing
mentoring at fitch
Fitch Ratings - AI/ML Team
Part of a 15-person AI/ML team, mentoring junior and senior engineers on AI/ML engineering best practices, system design, code quality, and production-readiness standards. Focused on building engineers who ship, not just prototype.
community mentoring
Mentoring & Community
Open to mentoring engineers and businesses navigating their AI journey - career transitions into ML, guiding a team through their first production AI initiative, or just thinking through a technical problem.
speaking & events
Conferences & Events
Keen to speak at conferences, tech meetups, startup events, and Python/JS community gatherings, especially in London. Happy to share what has been learned building AI in production.
contact
Open to senior IC roles, advisory, and consulting opportunities in AI/ML engineering.
New technical essays when they ship. No spam. Just the occasional note on AI systems and product engineering.