Engineering management that stays close to the architecture.
I'm a Chartered Engineer and engineering manager. Fifteen years across automotive systems, location data, media and fintech, leading teams that ship on event-driven, cloud-native platforms. I work through designs with senior engineers, question the limits and trade-offs, and set direction for the next level of scale.
At Zopa Bank, a UK-regulated bank, I lead the Auto Loans and Loan Decisioning teams on an event-driven lending platform built on Kafka. I drove the move to AI-native engineering across planning, build and delivery: feature-development allocation is three times what it was, with BAU reduced to match.
In my own time I built Leora, an identity-verified social-introduction platform, as sole engineer: 16 serverless, event-driven domains on AWS, native iOS and Android apps, and a production LLM agent with tested guardrails. It is how I stay current on what I ask my teams to build.
Experience
- Zopa BankEngineering Manager, Auto Loans and Loan Decisioning
- MonterosaEngineering Manager, 14 engineers across three teams
- GB GroupHead of Integrations
- Jaguar Land Rover and Tata MotorsSystems engineering, then Lead Engineer
- LeoraPersonal project, live on the App Store and Google Play
- Institution of Engineering and TechnologyChartered Engineer
What I bring
Event-driven architecture at scale
I have designed and led teams on event-driven platforms in three very different settings: a regulated bank, a multi-tenant SaaS platform under Black Friday load, and a serverless platform I built alone.
- Zopa Bank
- Lead two teams on an event-driven lending platform (Kafka) that mixes synchronous HTTP for immediate decisions with asynchronous events for everything downstream.
- GB Group
- Designed event-driven, stateless Node.js microservices for Loqate's multi-tenant Kubernetes platform, serving 5,000 HTTP requests per second during Black Friday with billing events emitted asynchronously off the request path.
- Leora
- 16 independently deployed domains communicating over EventBridge and SQS, Lambda consumers with dead-letter queues and idempotent handlers, a DynamoDB single-table design per domain, everything defined in CDK. Four months from conception to the App Stores.
AI-native engineering, measured
I treat AI as part of how a team engineers, not a tool bolted on. The standards come first, the output is held to them before review, and the result is measured.
- Zopa Bank
- Drove the Auto Loans and Decisioning teams onto AI-native engineering across planning, build and delivery. Feature-development allocation is three times what it was, with BAU reduced to match.
- Leora
- Architected and shipped an in-app AI agent: an OpenAI tool-calling agent reading live data across services, with grounding rules, intent classification that escalates to human moderators, allow-list URL sanitisation proven by adversarial tests, admin takeover, and per-plan quotas and token caps.
- Tooling
- Authored a Claude Code plugin of engineering standards, architecture patterns and 14 workflow commands (planning, plan audit, triage, bug fix, PR shepherding) so AI output meets standards before review.
Scaling teams and platforms
I have scaled teams through business growth, led platform migrations without downtime, and built the internal tooling that gives engineers their time back.
- Monterosa
- Defined an application architecture that let three teams (14 engineers) contribute in parallel while the business scaled. Cut maintenance 80% and brought new-engineer onboarding down to hours.
- Monterosa
- Led the migration of customers from legacy platforms onto the modern stack with no service disruption, and built Node.js CLI tooling that saved each developer three to four hours a week.
- Zopa Bank
- Led the Auto Loans team through significant turnover, building a structured knowledge-capture system so a largely new team could own the estate after its key knowledge holders left.
Leora
A personal project, and proof I hold myself to the standards I set for a team. Visit Leora
A serverless, event-driven platform, built in my own time as sole engineer.
- from conception to the App Stores
- 4 months
- independently deployed domains
- 16
- engineer, from design to production
- 1
Leora is an identity-verified social-introduction platform I designed and built end to end in my own time, to stay current on AWS and serverless design. Native iPhone (Swift) and Android (Kotlin) apps, a React and Next.js admin console, and a serverless, event-driven back end on AWS.
The back end is 16 independently deployable domains: profiles, matching, messaging, verification, moderation, billing, notifications and more. Domains never call each other synchronously. They publish events to EventBridge and consume them through SQS queues into Lambda handlers, so each one scales, deploys and fails independently. Consumers are idempotent and keyed on event id, dead-letter queues can be replayed, retries back off, and every moderation and verification decision lands in an immutable audit log.
Every domain is defined in CDK with its own pipeline, and promotion from development to production is gated on cross-service contract tests. AI wrote much of the code, held to standards I wrote first. The speed came from AI. The fact that it holds together came from fifteen years of systems engineering.
The in-app assistant is an OpenAI tool-calling agent that reads live data across domains, with grounding rules, intent classification that escalates to human moderators, adversarially tested URL sanitisation, admin takeover and per-plan token caps. Leora is live on the App Store and Google Play.
What I built it with
- Swift and SwiftUI
- Kotlin and Jetpack Compose
- React and Next.js admin console
- EventBridge, SQS and Lambda with dead-letter queues and idempotent consumers
- DynamoDB single-table design per domain
- API Gateway with Cognito auth, S3 and CloudFront
- AWS CDK in TypeScript, one pipeline per domain
- OpenAI tool-calling agent with tested guardrails
- Document OCR (Tesseract, OpenCV) and Whisper transcription
How I use AI
The rules I write for AI are the rules I set for a team: architecture first, standards written down, nothing shipped unproven.
Architecture first, then acceleration
I design the system: the services, the boundaries between them, the data and the ways it can fail. AI builds inside those lines, never across them.
Standards written before code
Every repository carries written rules the AI must follow, backed by shared standards for code quality, design and architecture. The first draft already matches the house style.
Nothing ships unproven
Typed code, automated tests and infrastructure as code in every service. Changes go to a development environment first, pass tests across services, and only then reach production.
Fifteen years of engineering
From cars to location data, fan engagement, media and fintech.
2011 to 2019
Jaguar Land Rover and Tata Motors
Systems engineering, then Lead Engineer
Led specification and systems design for a multi-modal speech recognition platform, spent five months in Bengaluru leading an offshore development team of 15 to 20 engineers, and progressed to Lead Engineer managing programmes and team members.
2019 to 2021
GB Group
Head of Integrations
Led nine engineers across two sub-teams, owning team structure, resource planning and hiring. Designed event-driven microservices for Loqate's multi-tenant, Kubernetes-based SaaS platform, handling 5,000 HTTP requests per second and emitting billing events asynchronously.
2021 to 2025
Monterosa
Engineering Manager, Front-End
Led three cross-functional remote teams (14 engineers) building and launching web and native applications while the business scaled. Defined the application architecture that let the teams contribute in parallel, led the migration of customers from legacy platforms with no service disruption, and built CLI tooling that saved each developer three to four hours a week.
2025
Tag
Senior Developer, Front-End AI Applications
Stepped back into hands-on engineering deliberately, to build production experience with AI platforms before leading teams that use them. Built the front-end application for a next-generation AI media generation platform.
2025 to now
Leora
Personal project
Designed and built the entire platform in my own time: native apps, an admin console and 16 serverless, event-driven domains on AWS. Live on the App Store and Google Play.
2026 to now
Zopa Bank
Engineering Manager, Auto Loans and Loan Decisioning
Lead two teams (8 engineers) owning lending decisioning and Auto Loans technology in a UK-regulated bank: hands-off on code, hands-on on technical direction and delivery. Drove the move to AI-native engineering across planning, build and delivery, and work with Zopa's AI teams on indexing engineering knowledge for LLM use.
About me
I trained as a systems engineer in automotive, at Jaguar Land Rover and Tata Motors, where a specification is a promise and every failure mode is thought through before anything is built. I brought that habit into software, building platforms and leading teams in location data, fan engagement, media and fintech. Today I lead the Auto Loans and Loan Decisioning teams at Zopa Bank: hands-off on code, hands-on on technical direction and delivery.
I stay close to the architecture. I would rather work through a design with a senior engineer, question its limits and trade-offs and set direction for the next level of scale than review it after the fact. I've scaled teams through business growth, led platform migrations without downtime, built internal tooling that gave developers hours back each week, and I'm comfortable talking through scaling trade-offs with engineers and with the people who fund them.
I've led remote and distributed teams for most of the last decade: 14 engineers across three remote teams at Monterosa, with developers working from Ghana, and an offshore team of 15 to 20 earlier in my career. Written standards, pinned-down specifications and problems surfaced early are how distributed teams work. Leora, a personal project, is where I hold myself to the same standards: one engineer, a complete production platform, built in my own time.
- Chartered Engineer, Institution of Engineering and Technology
- MEng (Hons) Electronic Engineering, University of Exeter
- AWS Certified Cloud Practitioner
- Speaker, AI Futures Day 2026, Manchester Digital Skills Festival
Get in touch.
If you are hiring an engineering manager, or you want to talk through event-driven systems, serverless on AWS or AI-native engineering, email me or find me on LinkedIn. I reply to everything.