Our vision

AI + datacenters are booming, electrification is accelerating, and trillions in generation + storage projects are queued — but the real bottleneck is the grid. Planning and operating networks is now a systems problem at national scale: complex models, messy data, strict governance, high-stakes decisions.

The next energy unicorn is the agentic fabric layer for grids — where truth lives, change is tracked, and humans + AI agents can safely operate together. If you want a Palantir-level challenge with real-world impact — and to help build the unicorn of power grids — Squid is it.

What is Squid 🦑

Squid is building that fabric: an agentic modelling platform that sits across utility teams and tools. We unify legacy planning models and operational data into a versioned, auditable source of truth, then layer on collaboration and AI agents that help teams ask better questions, run checks, explain changes, and move faster — without losing rigor.

We’re less than 3 months old and have one of the largest grid operators globally working with Squid. 

The role

You’ll be one of the first engineers building Squid day-to-day — high agency, high ownership, customer-facing. You’ll ship end-to-end (prototype → production → iterate), build core platform fundamentals, and create practical AI workflows .

We’re also AI-native in how we build: you’ll use tools like Cursor / Claude Code heavily, run agent swarms to build/QA/deploy, and design the infrastructure that lets us ship faster with high reliability. You’ll still do real engineering (backend, cloud, data systems) — but the default is “agent-first, human-in-the-loop.”

London in-person to start; will scale to more countries. 🦑

Requirements

We care more about taste, speed, and ownership than perfect keywords — but this is not a consumer app. You’ll probably thrive here if you:

• Have a genuine love for energy and power grids (you want to go deep)
• Are comfortable being customer-facing: talking to planners/operators, turning messy workflows into product, iterating fast
• Enjoy being full-stack — jumping across backend + frontend and doing whatever it takes to ship
• Like Palantir-style problems: complex data, permissions/auditability, workflows, reliability, explainability 
• Believe AI will change serious operational software, and want to build an AI company obsessed with grids
• Are excited to push AI tooling to it’s limits: you’ll be working and building a lot of AI infrastructure (prompts, evals, agents, harnesses, tests, tool APIs) alongside traditional code

Big pluses (either is great):

• You have strong power systems curiosity (even if you’re not a power engineer)

• You’ve built / operated agent infrastructure (evals, tool calling, guardrails, workflows, CI integration, swarm-style QA)

If you want to ship fast in a deep-tech domain, build agentic infrastructure, and work directly with the people operating the grid — Squid is definitely for you. 🦑