Services · Luke Parker, Senior Software Engineer

Hire me for the work with a deadline and a number attached.

I join your team as a senior engineer on React, Next.js and AWS serverless. I take a workstream from problem to production and leave it measured, tested and yours to run.

Available from 12 December 2026

Engagement
Contract, outside IR35
Location
Remote, UK working hours
Capacity
Full time
Minimum term
6 months, or a fixed scope build
Rate
On request

Inside IR35 considered for the right engagement.

What I offer

Six things I am hired to do

Checkout, conversion and performance

I work on the pages closest to the sale and measure what changes. After release, session data shows where shoppers stall, and the next release goes there.

  • React
  • Next.js
  • TypeScript
  • Core Web Vitals
  • Contentsquare

Proof

  • 3×

    Faster click & collect store list, with the store requests run in parallel.

    JD Sports
  • 1 page

    Checkout collapsed from several pages, with three further releases behind it.

    JD Sports

Full stack features inside your team

I join an existing team, work from its board and ship through its own review and release process, with unit and end to end tests on every feature.

  • React
  • Redux
  • Node.js
  • TypeScript
  • Jest
  • Playwright

Proof

  • 57

    Tickets delivered and 55 pull requests merged in six months.

    JD Sports
  • Paired on introducing TypeScript strict null checks on iPlayer Web, to keep null errors out of production.

    BBC iPlayer

AWS serverless and infrastructure as code

Event driven back ends defined in CDK, so every change is typed, reviewed and repeatable. The pipeline comes with it, and nothing is configured by hand in the console.

  • AWS CDK
  • Lambda
  • DynamoDB
  • EventBridge
  • SQS
  • Step Functions
  • CodePipeline
  • GitHub Actions

Proof

  • 299

    Automated tests behind a Magento and NetSuite sync on Lambda, SQS and Step Functions.

    Client middleware project, 2023
  • Lambda and EventBridge on least privilege IAM, defined in CDK and deployed through CodePipeline.

    BBC iPlayer

Content out of the codebase

When a content change needs a developer, a pull request and a deploy, I move it into the CMS and build the authoring tool around how the content team already works.

  • Amplience
  • Headless CMS
  • Content modelling
  • Data migration
  • TypeScript

Proof

  • 1,250+

    Size guides moved into Amplience. Sizing content now publishes with no developer, pull request or deploy.

    JD Sports
  • 13 markets

    Live on the new size guide, with 99.6% of products matched to the right guide before launch.

    JD Sports

AI assisted delivery and LLM features

Agent pipelines and agent driven review that sit inside the normal pull request process, with a human on the final step. LLM features on Bedrock or a vector store when the product needs one.

  • Claude
  • Amazon Bedrock
  • Pinecone
  • Playwright
  • Accessibility review

Proof

  • 46,000

    Files behind a support chatbot on a Pinecone vector store, with tiered access.

    Client chatbot project, 2023
  • An agent pipeline that backfills and corrects size charts, with promotion left to a person. Agent assisted accessibility review on pull requests.

    JD Sports

A product or internal tool, built from zero

For a direct client with a defined problem. Fixed scope, a working build to look at as it grows, and updates in plain English.

  • Next.js
  • TypeScript
  • Tailwind CSS
  • AWS Amplify
  • Vitest
  • Playwright

Proof

  • 8 panels

    Live data from eight sources on one display, delivered in a two week contract.

    Okto Technologies
  • 1,700

    Monthly players on Pass & Play Chess, my own product, with no marketing spend.

    Own product
One of Luke’s strengths is his willingness to challenge existing processes in a constructive way. He looks for opportunities to make things clearer, more refined, and easier to understand, not only for the current team but also for new members joining later.
Senior Software Engineer, BBC iPlayerLinkedIn recommendation

Process

How an engagement runs

  1. At the start

    Start with the outcome

    I ask what a piece of work is for before I pick it up, and we agree how it will be measured.

    At JD Sports

    The analytics event for the new size guide shipped ahead of the switch on, so old and new could be compared from a baseline.

  2. Every week

    Ship small and often

    Small changes behind pull request checks and a pipeline. You can read the board at any time.

    At JD Sports

    Each market had its own flag, created switched off. Any one market could be switched on alone and rolled back in about 15 minutes.

  3. At the end

    Leave it maintainable

    Tests and typed infrastructure as code. Your team keeps it going after I leave.

    At JD Sports

    I wrote the runbooks and the rollback plans, and led the knowledge transfer session.

Fit

Where I do my best work

  • A team that needs experience of how large organisations manage code and infrastructure
  • A team that wants a clear way to decide which changes come first
  • A workstream that needs one owner from problem to production
  • A rollout across markets that has to be staged and reversible
  • A system that needs mapping, to show its limits and where the opportunity lies
  • Ways of working that are ready to be questioned and refined
  • Workstreams and DevOps that need more from deterministic, automated checks and releases

Available from 12 December 2026.

Outside IR35, remote, six months or longer. Tell me the workstream and the date it has to land. I reply within 24 hours, and we start with a call.