Senior AI/ML Engineer

Company Description

Postura is a fast-growing deep-tech startup. We started two years ago as a team of four; today we are 10, and after recently closing a new investment round (to be announced soon), we are scaling fast and looking for new talent to double the team over the next year.

In manufacturing, improving workstations and production processes requires balancing worker safety and productivity through analyses that are still largely manual, time-intensive, and costly.

Starting from simple smartphone videos, our AI-powered software performs body-mechanics analysis and automates key ergonomic and industrial engineering standards (OCRA, NIOSH, MTM), replacing long observation times and manual assessments — delivering real-world AI built for accuracy and measurable impact.

We work with more than 20 major manufacturing groups, including Beko, Electrolux, and Piaggio, and are expanding across Europe with the goal of reaching 100 enterprise customers by 2026.

Role Description

As a Senior AI/ML Engineer, you’ll work on the core intelligence of our platform, at the intersection of computer vision, action recognition, and biomechanics, tackling real-world problems involving noisy data, complex human motion, and industrial constraints. This is work at the edge of what current applied AI can realistically deliver.

You will design and build systems that detect complex movements, gestures, and postures from video, enabling the platform to measure task execution times and ergonomic risks within a single, coherent model of human work.

You will join the AI team, working alongside two senior profiles in AI and computer vision, in close collaboration with the CTO. You’ll be fully involved in technical, product, and strategic decisions typical of a fast-growing startup.

What you’ll work on

• Improving the accuracy and robustness of our ergonomic and T&M AI systems on real industrial data.
• Designing Action Recognition models to identify tasks and measure their duration in production environments.
• Applying pose-tracking and geometric techniques to extract biomechanical and postural risk indicators.
• Translating ergonomic standards and T&M methodologies (e.g. OCRA, MTM) into scalable, automated systems.
• Bringing models end-to-end into production, ensuring performance, scalability, and reliability across multiple plants.
• End-to-end ownership. You own outcomes: from problem definition to real-world validation, iteration, and scale.

Tech Stack

• Languages: Python (core development)
• ML Frameworks: PyTorch, TensorFlow, Vertex AI
• APIs & Services: FastAPI, Pydantic
• Containerization: Docker
• GPU & Acceleration: NVIDIA GPUs, CUDA
• Cloud & Infrastructure: AWS and Google Cloud (GCP)
• Version Control: Git, GitHub
• Compute/Runtime: GPU-enabled environments for both training and inference deployments

What We Offer

• High-Impact, High-Expectation Environment. Your work directly shapes how companies improve ergonomics and productivity at scale. We value ownership, comfort with ambiguity, and measurable outcomes over safe, well-defined paths.
• Fast-Growing Startup Context. In a few months, you’ll gain responsibility and exposure that might take years elsewhere, as the team grows and our enterprise customer base expands across Italy and Europe.
• Flexible setup. A beautiful office in Milan or fully remote. We value outcomes and accountability over location.
• Team & Culture. We’re a young team (average age 29) that enjoys spending time together, both at work and outside of it. We value open, continuous feedback and stay flexible in how we work, so everyone feels comfortable and supported.

Requirements

• Get-Things-Done Mindset. This is a fast-growing startup environment. We’re looking for people who are eager to learn, willing to challenge themselves, and ready to go beyond the minimum when needed, to “spaccarsi il c**o”. It’s intense, high-ownership, and not for everyone, but also collaborative, rewarding, and fun.
• 5+ years of professional experience in AI/ML, with a strong focus on computer vision or motion analysis.
• Master’s degree or PhD in Computer Science, AI, Mathematics, or Engineering.
• Strong expertise in:

- Python and deep learning frameworks (PyTorch or TensorFlow);

- Computer vision for human movement, including pose estimation, action recognition, and temporal

modeling;

- 3D geometry, linear algebra,and statistics applied to biomechanical or motion-related problems;

• Proven experience building, shipping, and maintaining production AI systems, including GPU-based training/inference, Git-based workflows, and robust validation on real-world data.
• Experience using AI-assisted, spec-driven development workflows (e.g. Cursor, OpenSpec, or equivalent).
• Experience integrating modern AI model APIs (e.g. GPT, Gemini or similar) into real systems, including prompt design or model adaptation.
• Fluent English (written and spoken).

Extra Skills

• Background in biomechanics or human movement analysis.
• Familiarity with industrial engineering or Time & Methods standards.
• Experience with Edge AI or deploying models under real-world constraints (latency, compute limits, noisy inputs).
• Experience with Agile Methodologies: Previous exposure to Agile/Scrum workflows is appreciated to better align with our iterative delivery process.
• Previous experience in early-stage or high-intensity startup environments, with a strong sense of ownership and adaptability.

Diversity and Inclusion

We value diverse perspectives and backgrounds. If you don’t meet every requirement but believe you can make an impact, we encourage you to apply.

Location

Milan (Italy) or fully remote.

If you do not have an Italian VAT number (Partita IVA) and do not reside in Italy, you may be required to relocate your fiscal residence to Italy. We’ll support you in understanding relocation options and available tax benefits.

Hiring Process

• Screening call (~20min)
• Technical interview (~45 min)
• Interview with the founding team (~20min)

⚠ We keep the hiring process lean and adaptable. It may vary slightly

We move fast, but we’re a small team building hard things. If we’re slow to reply or can’t respond to every application, we’re probably busy building. Thanks for understanding.