Founding Backend Engineer

Why this company, why now


Energy is the binding constraint on AI and economic prosperity: nothing scales until it's cheap and abundant and sustainable. Sustainable energy is non-negotiable during a worsening climate crisis, but the forces driving the energy transition are now also: resilience, sovereignty and national security.

SOLR AI builds at that pressure point: Frontier intelligence for the energy transition.

Renewable energy intelligence.

The renewable energy industry still runs on spreadsheets, PDFs and weeks of manual survey work.

SOLR AI is the intelligence layer replacing it: computer vision and machine learning over aerial imagery, geospatial, property, grid and weather data and more, turning a UK address into a bankable-grade renewable energy assessment in minutes rather than weeks. The UK is the first market, not the last.

About the company: Stanley Wilson, co-founder and CEO, started building the company before ChatGPT was released;
Dr Anna Chabokdast, co-founder and CTO (ex tractable), trained her first ML model ten years ago. We were building at the frontier of AI in energy before the hype.

Since then: our own models trained and in production, national coverage across all 14 UK distribution network regions, commercial energy operators live on the platform, first enterprise contracts converting, and the next generation of our models headed for the UK's fastest AI supercomputer. VC-backed and supported by Google for Startups, Barclays Eagle Labs and Sustainable Ventures.

How we operate

• Frontier science, shipped. We apply the frontier of AI research to real energy problems, and we're judged by what reaches production.
• Engineers are the fabric. Engineers own problems end to end: define, build, talk to users, ship. There is no layer between your work and the outcome.
• Extreme ownership, high slope. We hire for agency, learning rate and humility over pedigree. You own the outcome, including what you get wrong.
• Research-led product development, product-led growth. Every research bet is measured by customer value; the product moulds our go-to-market.
• Clarity and candour. We write things down, build in the open, and give direct feedback.
• Intensity, honestly. This is seriously hard work at the edge of what's known. It's not for everyone, and that's the point.

Requirements

The role

You'll work directly with Anna, our co-founder and CTO, as our first founding engineer. The near-term job is turning a platform that works into a platform we can put contracts and SLAs behind: reliable, observable, secure by default, and able to onboard paying enterprise clients without every integration being a fire drill. Most of the day-to-day is backend and data engineering, and the scope is founding-level: you'll shape the architecture with Anna, own whole systems end to end, and touch whatever the platform needs.

Over time the role grows into deploying and operating the production ML that powers the pipeline alongside Anna, so we'll want proof you've already taken a deep learning model into production.

The hard problem underneath all of it: standing behind accuracy claims contractually, on top of messy, heterogeneous external data, at national scale. In your first 90 days you'd ship multi-tenant authentication, a hardened ingestion layer with the observability we can put an SLA behind, and the benchmarking harness for those accuracy claims. Real production milestones, not onboarding theatre.

On AI tooling and ownership

Daily, fluent use of AI coding and agent tools is non-negotiable in this role, and the work trial will assess how you work with them. But we've seen the common failure mode where the agent quietly becomes the owner of the codebase.

We want the opposite: you use these tools to move fast, and you can explain, defend and rebuild any part of the system without them. If you can't say why a piece of code exists, it doesn't ship.

What you'll work on

• Harden the core pipeline against failure: timeouts, retries and graceful degradation across external data dependencies
• Build out CI/CD, structured logging and alerting so we know about failures before clients do
• Design and ship multi-tenant API authentication for design partners and enterprise clients
• Own data engineering across the pipeline: ingestion, schema design and geospatial queries over property, building and imagery data
• Build the automated testing and ground-truth benchmarking that lets us stand behind our accuracy claims
• Work directly with our design partners and enterprise clients during onboarding and integration, and turn what you hear into what we build next
• Shape architecture decisions directly with Anna as we onboard clients and scale, building to the standard that enterprise security and data-protection reviews expect, by default rather than retrofit
  

What we're looking for

• Strong backend fundamentals: you can evaluate different ways of building an API and choose the right one for the problem, rather than defaulting to a framework
• You've shipped and operated a real product in production, not just built one: you've carried on-call, handled an incident, or owned uptime for something people depended on
• Data engineering judgement: pipelines over external APIs and data feeds, and the sense to pick the right schema and database for the problem
• You've taken a deep learning model from prototype to production deployment
• Engineering discipline as habit, not policy: testing, code review, secrets management, sane error handling, cloud-native by default
  

Nice to have: geospatial or GIS work; experience in compliance-conscious environments (GDPR-heavy, fintech, healthtech); DevOps exposure; enough frontend competence to extend a client-facing interface; early-stage startup background.

Who shouldn't join

If you want a spec handed to you, a team to disappear into, this isn't it: the near-term job is reliability engineering with a contract riding on it. If the uncertainty in what the company looks like in a year drives fear rather than excitement, then this isn't a fit.

Stack

Python backend, Typescript, a Postgres-family database with geospatial support, cloud-native infrastructure on GCP, and a modern frontend framework for the client-facing interface. We'll go deep on the architecture with you during the process.

Benefits

Compensation and equity, plainly

£85-95K base plus 1.25-1.75% of the company as EMI options (six-year vest, one-year cliff). We assess hard and we pay for demonstrated competence: the process maps you to a level, and each level carries a fixed salary and equity number at the top of what we'd pay for it. You get our best offer first, so you never have to negotiate for it.

Before you sign we'll walk you through the real numbers: the strike price, the current valuation, what dilution across future rounds realistically does, and the EMI tax treatment that makes this the most efficient equity a UK employee can hold. Plus pension and other benefits.