← Back to nathanatherton.comDownload PDF

Nathan Atherton

Staff Software Engineer · UK / Remote

nathanatherton2@gmail.com · nathanatherton.com · linkedin.com/in/nathan-atherton · github.com/nathanavie

Summary

Staff Software Engineer shipping AI-first products at scale, building since 2019. Day-to-day work spans agent systems in production, platform engineering, and developer experience on a learning platform serving 5M+ learners across 60+ countries. Promoted twice in under five years at Thrive (Senior → Lead → Staff). Also ships personal products; most recently RL Track, a Discord bot with ~1.4M registered users, ~3K daily active, across 11K+ Discord servers.

I work best in small, async-friendly teams that sit close to the business. I mentor naturally and tend to be the person teammates pull in when something is on fire or when a new way of working needs to land. UK-based and fully remote by default; I've twice shipped major work while travelling across Europe (two months in 2023 and an extended stint through 2026).

Experience

Thrive (Jun 2021 to Present)

A next-generation learning platform reaching 5M+ learners across 60+ countries.

Staff Software Engineer · Feb 2026 to Present

Lead Frontend Engineer · Sep 2024 to Jan 2026

Senior Frontend Engineer · Jun 2021 to Aug 2024

  • Kiki, the in-product AI agent: co-built the original Kiki prototype at a Thrive hackathon with two others; shipped into main within a month and has since grown into its own dedicated team and product. Now lead engineer on it. On rejoining the team I led a focused 30-day push that shipped a major UI/UX overhaul and several new interaction modes, including a content-coaching mode and an analyst mode surfacing live Snowflake-backed analytics as inline charts. Kiki lets learners ask questions about any content, get coached on topics, and sign up to events; moving the platform toward AI-as-an-interface.
  • AI Content Creator: shipped the AI authoring workflow now used as the primary way content is created on Thrive (text, image, quiz, rewrite, course-from-PDF). Currently extending with the Head of Engineering to generate SCORM packages directly from AI, a capability sales confirms no competitor ships.
  • Google social feed: core engineer on the team that built and shipped an in-app social feed for Google, launched in Japan and then the US. Delivered in 6 months; Google's team noted prior vendors had scoped comparable work at ~3 years.
  • Trusted for high-stakes delivery crunches: including direct CPTO-requested work and fixes during the final push that helped land the Emirates enterprise deal.
  • Internal AI Slack bot: built a Cursor-CLI-orchestrated bot that fans out to sub-agents across selected repos to answer engineering questions in Slack. Served 600+ questions in 3 months, unlocking significant engineering time previously lost to ad-hoc Q&A.
  • CI/CD overhaul: migrated the frontend Nx monorepo from Bitbucket to GitHub Actions, cutting the PR pipeline from ~1 hour to under 7 minutes.
  • Build system overhaul: replaced Webpack with Rspack across the monorepo. Local builds 4 to 5 min down to under 1 min, CI deploys 15 min down to ~3 min.
  • DevEx automation: built custom GitHub Actions for Jira-readiness checks, merge gating on column state, third-party platform key validation, and AI code-review tooling, plus an interactive SCORM Git-basics course (generated end-to-end with Claude) now required before non-engineer GitHub access.
  • People & leadership (during Lead tenure): mentored two Senior engineers who were both promoted to Lead during my time as their manager; recognised by the VP of Engineering as the most proactive manager at Thrive in that period. Stepped back to IC on Staff promotion to focus on cross-cutting product and platform work.
  • Delivery turnarounds: brought at-risk teams back on track, repeatedly taking teams from ~2 milestones behind to on-target delivery in under a month.
  • Scaling through acquisitions: go-to onboarding point of contact for engineers joining through multiple company acquisitions, and conducted technical interviews across frontend, full-stack and Lead roles.
  • Architectural authority & internal influence: go-to engineer for structural decisions across the monorepo (library extraction, dependency boundaries, legacy retirement, microfrontend patterns) as one of its founding contributors. Demoed the AI-first development workflow at a company-wide AI Day; shared the setup gist that engineers across the company now use day-to-day.

Now Signage (Mar 2021 to Jun 2021)

Sole frontend engineer on a digital-signage product running across mixed hardware. React, Electron, Docker.

CDL Software (Jun 2019 to Mar 2021)

Built insurance quote journeys (house, car, pet) in agile delivery teams. React, TypeScript, Java, Cypress.

Skills

AI & Agents · Multi-agent orchestration, MCP (Model Context Protocol), tool use, prompt engineering, RAG, Claude Code, Claude Agent SDK, LLM integration, AI code-review patterns

Frontend · React, TypeScript, JavaScript, Next.js, Tailwind, shadcn/ui, Framer Motion, Three.js / React Three Fiber, Vite, Rspack, Apollo GraphQL, Storybook

Backend & Infrastructure · Node.js, Bun, GraphQL, PostgreSQL, SQLite, Docker, Netlify, Cloudflare Workers, GitHub Actions

Practices · System design, test-driven development, performance optimisation, observability, developer experience, technical mentoring, microfrontend architecture, monorepo tooling (Nx)

Side Projects

RL Track · rltrack.netlify.app · Discord bot serving live Rocket League rank cards via a server-rendered canvas pipeline. ~1.4M registered users, ~3K daily active, across 11K+ Discord servers since 2021.

nathanatherton.com · nathanatherton.com · My personal portfolio and blog. Built AI-first with multi-agent orchestration, fully prerendered, with a 3D interactive easter-egg /about experience. Hosts 20+ posts on AI-first development, side projects, and staff-level engineering.

Monitoring the Situation · A geopolitical event monitor I built mid-travel during the 2026 USA/Iran tension when the Strait of Hormuz was closing and onward flights were at risk. Polls intelligence sources on a schedule, summarises with an LLM, and pushes alerts when the situation materially changes. In daily personal use.

Writing

Recent posts at nathanatherton.com/blog:

  • 1.4 Million Users and a Raspberry Pi at My Parents' House
  • Staff Engineering in the Age of AI: What Actually Changes
  • Asking Claude What's on My Driveway

Education & Interests

Edge Hill University · Sep 2016 to Jul 2019 · 1st Class BSc (Hons) Web Design and Development

Outside work: football, video games, animals, AI tinkering, and travelling (with a laptop).