Founding Full Stack Engineer - Computational Geometry in London

Location: London, United Kingdom (Hybrid)

Salary: Actually competitive. We want the top 0.1% and we’re willing to pay for it.

Equity: Meaningful founding stake—Bench is building Engineering Superintelligence.

This isn’t a job for everyone. We’re after the top 0.1% of mission-driven builders. More than anything, we care about hunger, passion for this problem space and your track record of exceptional achievement.

We’re building a category-defining company from first principles. The kind of product that will reshape how engineering teams work - and a journey you’ll look back on as the most intense, creative, and career-defining chapter of your life.

It’ll be hard. You’ll move fast, take ownership, and ship real things that matter - alongside a small team of obsessive builders who hold themselves to a high bar.

—Bench - The Mission

Building hardware is still painfully slow - repetitive workflows, disconnected tools, and endless manual iteration. We’re building the platform that changes that. Bench gives hardware engineers what software teams have had for years: a fast, flexible, AI-native stack that automates the boring parts and supercharges the creative ones.

Bench is a collaborative system where AI agents help engineers plan, build, and run workflows securely and at scale. But for an AI to actually "engineer," it first needs to understand the physical world.

The mission is simple: make hardware development 10x faster - and unlock the next generation of physical innovation. This is a generational opportunity. We’re well funded, agile, and shipping fast - and the engineers who join now will be able to truly shape our trajectory.

—The Role

Focus: Engineering Spatial Intelligence. Current AI models are brilliant at text but illiterate in 3D space. You will fix that.

This is a deep tech role for a builder who loves hard math - you will own the high-performance C++ core that serves as the geometric intuition for our ingress and planning AI stacks. Your job is to bridge the fundamental disconnect between raw geometric data (B-Reps, surfaces, topology) and high-level reasoning.

One day, you might be optimising an intersection algorithm to run in microseconds; the next, you might be exposing those geometric insights to our AI stacks or shipping a new visualisation tool to production. This will be a very versatile and flexible role.

What You’ll Own

• Architecting and building the high-performance C++ engine that serves as the foundation of our IRIS ingestion pipeline.
• Developing novel algorithms to extract "design intent" from dumb geometry, turning raw CAD data into semantic context that our agents can reason about.
• Ensuring our core pipeline handles complex, messy, real-world engineering data with near-instant speed.
• Working directly with our AI engineers to feed your geometric insights into our models.

Requirements

What we care about the most is that you’re hungry, sharp, and have a record of exceptional output. Other than that:

• We only hire engineers. You need an engineering degree (Mechanical, Aerospace, or Mechatronics/Robotics preferred) or have actual hands-on experience building hardware. We want you to have lived the problem that we’re solving.
• Mathematical Fluency. You are comfortable with linear algebra, 3D vector math, and topology.
• 1-3 YOE in a fast-paced startup, deep tech, or high-intensity environment.
• A First-Principles Mindset and are excited to architect a system from scratch to solve a problem that hasn't been solved before.
• Extreme Ownership. You’re happy to work with any part of the stack, from the kernel to our agent frameworks and cloud.
• You're an AI-native developer - you already use Cursor, Claude Code, or similar tools to brainstorm, code, and iterate faster, and you're eager to push these to their limits.
• Bonus: Experience building or working deeply with geometry kernels, physics engines, or rendering pipelines.
• Previous founding experience. Whether a side hustle, a university society, or a full-fledged company.

Why Join Us

• Unique industry opportunity - We’re one of the handful of companies in Europe building AI-native software for engineering - a category that will underpin the next wave of physical innovation. Joining now means helping define how this industry evolves.
• True "Hard Tech" - This isn't another CRUD app! You are building the foundations for a new category of software, solving problems at the intersection of geometry, graph theory, and AI.
• Uncapped learning curve - No other role will give you this breadth of technical challenge this early in your career.
• Meaningful equity - As an early team member, you’ll have significant participation in the upside.
• The people - Work with obsessive builders who breathe and sleep reshaping how hardware gets made.

The Founders & Team

In this role, you’ll work very closely with both of the founders - you’ll be reporting to Raihaan.

Raihaan Usman, CTO - Aerospace Engineer, previously Engineering Lead at Imperial’s Karman Space Programme, an academic team of 50+ engineers aiming to be the first academic team to launch a reusable rocket into space. Shipped software patches to a live satellite at ESA.

Martin Bielicki, CEO - Mechanical Engineer, previously founded and led Hyperlink, a team of 100+ engineers that designed and built London’s first Hyperloop Pod. Founded companies in e-commerce and digital marketing and worked in venture building.

Alongside Martin and Raihaan, you’ll be joining two cracked founding engineers, both Aerospace Engineering graduates from Imperial College London who have built at the cutting edge of Rolls-Royce, Mathworks, Autodesk, and top AI startups.

Compensation

Competitive (but actually - we want the top 0.1% and we’re willing to pay for it). Meaningful equity stake.

The Interview Process

• Submit your CV and answer the 2 application questions.
• 30 Minute Screening Call with Martin.
• Short take-home technical task.
• 60-minute technical interview with Raihaan.
• Last stage, office half-day / interview with Martin and Raihaan.