T
Tractable Tech Stack
Computer vision AI for damage assessment and property inspection
Tractable applies deep learning to visual damage assessment in insurance and automotive — using Python, PyTorch, TensorFlow, and a mix of cloud (AWS, GCP, Azure) plus edge compute (Raspberry Pi, NVIDIA Jetson). The tech stack signals a company moving computer vision from research into production: heavy investment in model frameworks paired with containerization (Docker, Kubernetes) and edge hardware reveals the engineering challenge of deploying vision models at scale across field devices. Active hiring leans senior (80% of roles), focused on engineering and data, indicating a team scaling ML systems and infrastructure rather than headcount broadly.
Tech Stack 28 technologies
Core StackPython PyTorch TensorFlow AWS Linux TypeScript PostgreSQL GitHub Docker Kubernetes Rust Go Windows Git Remix GCP Azure Raspberry Pi NVIDIA Jetson Arduino RS-485 CAN Modbus UART I2C GPIO C/C++ LiDAR
What Tractable Is Building
◆Challenges
- Scaling ml infrastructure
- Performance challenges
- Support complex workflows
- Scaling financial operations
- Internal control environment
- Financial compliance across regions
- Assembly process inefficiencies
- Scaling ml solutions in production
- Enhancing performance and scalability
▲Active Projects
- Scalable ml systems for computer vision
- Ai solutions for insurance and automotive
- Deep learning model productionisation
- Ml & data platform
- Ml lifecycle optimization
- Productionising deep learning models using python and modern ml frameworks
- Designing ai interaction models for complex workflows
- Drive-through imaging rig
- Seamless integration of ai solutions
- Luma
Hiring Activity
Steady10 roles · 3 in 30d
Department
Seniority
Company intelligence
