James Phoenix

AI Infrastructure Engineer

James Phoenix

I build AI systems that work in production, not demos that break under load.

350K+
latest project code size
306K+
developers taught
10K+
O'Reilly books sold
45+
bootcamps delivered

Services

What I Build

Production-grade systems with observability, reliability, and performance at scale.

  • SaaS Applications

    Full-stack SaaS development from idea to production. Including auth, payments, analytics, and automated testing.

    Learn more
  • LLM Engineering

    Building custom AI solutions and integrating LLMs into existing applications. Specializing in RAG, fine-tuning, and prompt engineering.

    Learn more
  • React Development

    Modern web applications built with Next.js, TypeScript, and Tailwind. Focus on performance, accessibility, and user experience.

    Learn more
  • Python Development

    Building scalable backend services, data pipelines, and automation tools. Expert in FastAPI, Django, and async Python.

    Learn more
  • Data Engineering

    End-to-end data pipeline development, ETL processes, and data warehouse solutions. Expertise in real-time and batch processing.

    Learn more
  • Web Scraping

    Building reliable and scalable web scrapers with respect for robots.txt. Expert in handling JavaScript-heavy sites and CAPTCHAs.

    Learn more

Custom AI Development

Agentic Software Development

I build custom infrastructure that helps development teams ship faster with coding agents. The right testing and tooling setup can multiply your team's output while keeping humans on the loop.

What I Build

  • Custom Agents
    AI agents trained on your specific codebase and coding conventions
  • Testing Infrastructure for Agents
    Infrastructure that allows your development team to scale with coding agents, whilst having a human on the loop
  • Repository Engineering
    Improving your developer environment and codebase structure to maximize coding agent performance
  • Production Infrastructure
    Observability, rate limiting, and multi-provider failover
~/project
$claude "Build auth system with OAuth and tests"
Createdsrc/auth/oauth.ts
Createdsrc/auth/middleware.ts
Createdtests/auth.test.ts
Createdtests/oauth.test.ts
✓ All 18 tests passing
Coverage: 94% | Time: 1.8s
$git commit -m "Add OAuth authentication"

Technologies

Battle-tested stack

From frontend to infrastructure, every tool chosen for production reliability. These aren't just technologies I know. They're technologies I've deployed, scaled, and debugged at 3am.

Full-stack TypeScript

End-to-end type safety from React to API

Production observability

Traces, metrics, and logs with OpenTelemetry

Infrastructure as code

Reproducible deployments with Terraform

TypeScript iconTypeScript
React iconReact
NextJS iconNext.js
Python iconPython
Node.js iconNode.js
PostgreSQL iconPostgreSQL
Redis iconRedis
Docker iconDocker
Google Cloud iconGCP
Terraform iconTerraform
OpenTelemetry iconOpenTelemetry
Prometheus iconPrometheus
Grafana iconGrafana

Client Reviews

What Clients Say

More reviews on Google
Rhys Fisher
Rhys Fisher
James has been a key measurement partner for a number of years. While running a boutique analytics agency, I would bring Just Understanding Data into a number of growth transformation projects at a time, across a portfolio of B2B SaaS and D2C E-commerce clients. These projects involved solving a diverse set of challenges. Together we worked on bespoke data collection solutions, applying machine learning models to marketing challenges, and surfacing often complex ideas into easy to digest actionables that clients could act on. James is well spoken and handles client communications well. His super power is being able to create a shared understanding around the insights drawn with analytics. Without James, you’ll be leaving growth on the table. 5* recommendation from me.
Shuaib Ahmed
Shuaib Ahmed
James is literally a legend! I asked him to perform a web scraping task that was quite tedious revolving academic abstracts and he literally did it within a week after posting the gig. He also went above and beyond and gave me tips on applying NLP techniques which worked out quite well for my solution the month after. I would highly recommend working with him for any of your data collection needs or if you just need tips/help solving any Data Science issues!
Jeanne Keith Felipe
Jeanne Keith Felipe
James is a very nice and professional person to do business with. He doesn't mind explaining technical information to give you a clearer picture of how online business works. He has a great work ethic and I'm sure Insight Jar will go far

Ready to build something great?

Production-grade software and data engineering. Let's discuss your project.

Book a Call

Blog

Latest Articles

Thought Leaders thumbnail

Thought Leaders

Read article
Systems Thinking & Observability thumbnail

Systems Thinking & Observability

Read article
Career Archetypes thumbnail

Career Archetypes

Read article