Currently architecting GenAI for a global aerospace engine manufacturer

I design generative AI systems that people use every day.

I work with large enterprises and own each system from PoC and proposal through to production.

See the work Contact

<90 sto find relevant emails in a 1M+ archive
<2 minto generate a project charter that took 30–40 effort hours
300+users on a production GenAI platform I architected
8AI and data solutions grown from one proof of concept

Delivered in

ManufacturingAerospace, Steel
Energy & UtilitiesElectricity, Gas
TravelCorporate, Leisure

For clients across the UK, Canada and Australia.

Selected work

Support engineers search 15+ years of engineering email in plain English.

Featured · Global aerospace engine manufacturer

AOG engineering knowledge assistant

Problem

The 24/7 Aircraft-on-Ground (AOG) support team relied on more than a million engineering emails, but finding the right one meant 30+ minutes of repeated mailbox searching.

Approach

Engineers ask a question in plain English and can put a keyword in "double quotes" to make sure it appears in every result. They can switch between keyword, semantic and hybrid search. Matches come from semantic and BM25 search over a Lakebase index, reranked with Cohere Rerank, and each email is shown with an AI summary and its source details. Engineers can ask follow-up questions, LLM guardrails sit in front of the model, and new emails are added daily by a medallion pipeline on Databricks.

Outcome

Engineers ask once and get the right email, instead of running search after search in Outlook. The assistant was validated in staged UAT and released to production.

My role: AI Solutions Architect
Delivery: Team of 5
Status: In production

Read the case study

AI project initiation & charter platform
Production
An AI project initiation platform that project managers use to launch new projects, from guided intake to a first-draft charter, with 300+ users onboarded.
LangChain · Azure OpenAI · AKS

See all work

Capabilities

Architecture that survives security review and real users.

Agentic & GenAI architecture
LangGraph and LangChain workflows, Azure OpenAI, multimodal and vision LLMs, prompt and context design.
Retrieval engineering
Hybrid vector + BM25 search, chunking and reranking, with search modes users can switch between, over a million documents and up.
Enterprise hardening
Guardrails, access control, SonarQube quality gates, observability and MLflow‑based LLMOps on AKS.
Data science foundations
Causal inference with DoWhy, change‑point detection, statistical process control, classical ML.

I design generative AI systems that people use every day.

Support engineers search 15+ years of engineering email in plain English.

AOG engineering knowledge assistant

Problem

Approach

Outcome

AI project initiation & charter platform

Architecture that survives security review and real users.

Agentic & GenAI architecture

Retrieval engineering

Enterprise hardening

Data science foundations