Robotics Engineer

Cota Robotics; Stockholms kommun; On-site; SEK 700,000 - 1,000,000; Full time

Job summary

Cota Robotics is a software-focused robotics startup that places robots in live stores to perform shelves and pallets work. You’ll develop and deploy manipulation and navigation policies on real robots, collaborating with hardware teams as needed to ensure reliable operation in stores at dawn.

You’ll balance learned control with classical planning and perception, deciding which approach fits each problem and taking ownership from store observation to shipped behavior in a fast, chaotic

Qualifications

• Experience getting software to run on real robots, not only simulation.
• Understanding when to use classical vs learned approaches.
• Ability to define problems before solving them.
• Comfort with fast-paced, chaotic environments.

Responsibilities

• Develop and deploy manipulation and navigation policies on real robots in real stores.
• Work with VLAs and modern learned control alongside classical planning, perception, and control.
• Decide which approach fits which problem, and be right about it often enough.
• Own problems from store observation to shipped behavior.

Skills

Robotics software integration

Real robots deployment

System architecture tradeoffs

Fast learning & problem framing

Job description

Cota Robotics puts robots to work in retail stores. They take on the repetitive, time‑consuming jobs like inbounding goods, replenishing shelves and sorting stock, currently mostly overnight. They're driven from our control room in Stockholm and increasingly run on their own. We raised a €1.92M pre‑seed in August 2026 and are live with retailers in Sweden and Finland. We're a small team based at Sankt Eriksgatan in Stockholm, where half the office is our own dark store.

About the role

You’ll build the autonomy that lets our robots work shelves and pallets in live stores. This is a software role: you’ll work closely with hardware, but you won’t design it. We care about what works in a store at 6am, not what wins a benchmark, and that usually means combining learned policies with classical methods rather than picking a side.

What you’ll do

• Develop and deploy manipulation and navigation policies on real robots in real stores
• Work with VLAs and modern learned control alongside classical planning, perception, and control
• Decide which approach fits which problem, and be right about it often enough
• Own problems from store observation to shipped behavior

What we’re looking for

• Hands‑on experience getting software to run on real robots, not only in simulation
• Clear on the tradeoffs between classical and learned approaches
• A fast learner who can define the problem before solving it
• Comfortable with chaos and speed
• We expect to fill this role in November or December 2026
• You must already have the right to work in Sweden, as unfortunately, we’re not able to offer visa sponsorship at our current stage