I'm Kam Cheng. I build reliable software, from Python services and APIs to the interfaces people use, and I've led a distributed engineering team across four continents.
Fig. 1 Approach, one engagementSchematic, not to scale
Diagram of one engagement. A piece of work, drawn as a sheet, moves along a line through four stages: Discovery, Proposal, Delivery and Handover. At each stage a checkpoint is agreed with you, and the stages are marked agreed, signed off, shipped and handed over in turn, while a progress bar fills to the end of each finished stage.
Services
Engagements run through KC Consulting Ltd
01
Backend and API services
Python services, APIs and integrations on Google Cloud, built to handle volume and fail safely.
Extending and hardening SLA-bound batch pipelines with audit and validation gates, orchestration and monitoring.
Airflow, Trino/Starburst, SQLMesh
03
Technical leadership
Interim or fractional team leadership, covering roadmap, hiring, code review, on-call and incident practice.
1:1s, sprint planning, post-mortems
04
AI-ready engineering
Helping teams adopt AI coding tools safely: plan-first and test-first workflows with automated and human review, and an honest look at how ready your codebase and delivery practice are.
Diagram of one message’s failure path. Requests arrive over HTTP into Pub/Sub, the service pulls them, and outbound messages are sent. One message keeps failing and is retried, up to 10 times. When the last retry also fails, the message, drawn hollow, drops to the dead-letter queue, and the queue raises a developer alert. Schematic, not to scale.
One client's contacts page went from 84 seconds to under one second, and heavy reporting queries ran up to 22 times faster. I set the approach and directed the work.
A short call to understand the problem, the constraints and the outcome you need.
02
Proposal
A written scope with milestones, risks and an agreed point of review before work begins.
03
Delivery
Test-first engineering in short iterations, with automated and human code review and regular demonstrations of working software.
04
Handover
Documentation, runbooks and monitoring in place, so your team can run and extend what was built.
About
Software engineer and engineering lead
Fig. 8 Kam Cheng · Engineering lead
Before software I spent ten years in insurance fraud investigation and legal claims, which gave me a habit of following the evidence and getting the details right.
I moved into engineering in 2022. Since then I've worked on healthcare data pipelines, built messaging services, and led an engineering team for a business-messaging platform.
Fig. 9 Career2011 to 2026
Insurance fraud and legal claimsDWF, AXA, RSA2011 to 2021
Ten years investigating suspected fraud and handling legal claims at DWF, AXA and RSA, including a caseload of around 120 complex claims at a time.
Harvard CS50x2021
Northcoders bootcamp2022
Harvard's CS50x online in 2021, then the Northcoders full-stack bootcamp in 2022: algorithms and data structures, then JavaScript, React, Node.js, PostgreSQL and test-driven development.
TwilioSoftware Engineer Apprentice2022
An apprenticeship in the Microvisor IoT team, where I built an automated crash-reporting service that sent engineers a daily summary of kernel failures.
KC Consulting Ltd · 2022 to 2023
Web Developere-commerce startup2022 to 2023
Front-end work for an e-commerce startup: the React interface for adding products and selected storefront pages for its Shopify marketplace.
Designed and built the production Pub/Sub messaging service, load-tested it to calibrate its autoscaling, and wrote best-practice documentation that went into the client's official partner documentation.
EMISData Engineer2023 to 2025
Helped build, maintain and extend healthcare data pipelines on Apache Airflow, Trino and AWS against a 24-hour SLA, adding audit and validation steps before data reached the curated tables.
KC Consulting Ltd · 2025 to 2026
Team Leadbusiness-messaging platform2025 to 2026
Led five engineers across four continents, covering sprint planning, hiring, on-call and post-mortems. I stayed hands-on too: I built a self-service reporting tool, led a cross-service migration and directed a database optimisation.
