Join a leading AI lab to build safety-aware autonomous systems focusing on software engineering, data science, and AI research.

About the Company

Lexsi Labs is a pioneering AI research lab dedicated to developing aligned, interpretable, and safe superintelligent systems. Based in Mumbai, with presence in Paris and London, our lab has published over 25 research papers in prominent conferences such as ICLR, ICML, and EurIPS in the last 15 months. With a commitment to high autonomy and a flat organizational structure, we empower engineers to take full ownership of their work, from conception to deployment.

Role Overview

We are seeking a Software Engineer to contribute to our mission of building safety-aware autonomous systems. This role will require working on our current project aimed at enhancing the harness, execution substrate, and evaluation systems for our agents, ensuring that they perform optimally within customer environments. You will be instrumental in creating the foundational layers that support our agents across software engineering, AI research, and data science teams.

Key Responsibilities

• Harness Development: Design the core agent loop and execution model, focusing on orchestration, concurrency, and state management. Implement failure semantic policies to enhance robustness.
• Evaluation Systems: Develop task suites for various agent types and infrastructure to run them efficiently. Create verifiers and scoring systems to ensure quality and performance.
• Cross-Team Collaboration: Work closely with different agent teams and the research department to align evaluation designs, interpretability, and post-training assessments of agent behavior.

Tech Stack

• Programming Languages: Python (advanced level)
• Tools and Environments: Docker, OCI
• Systems: Distributed systems, concurrency handling
• Testing Frameworks: CI/CD tools, automated testing harnesses
• Monitoring: Distributed tracing, structured logging
• Algorithms: Statistical analysis and measurement techniques

This position provides a unique opportunity to work at the cutting edge of AI technology and to play a crucial role in the development of systems that prioritize safety and performance.

• Strong software engineering fundamentals and advanced proficiency in Python.
• Experience in concurrency and distributed systems, including async execution and worker pools.
• Knowledge of containers and sandboxing, particularly with Docker.
• Familiarity with test and CI infrastructure, building test harnesses, and running parallel testing suites.
• Proficiency in measurement literacy, statistical analysis, and observability practices like distributed tracing.
• Strong debugging skills and the ability to read between the lines of loosely specified problems.
• Previous experience with evaluation or benchmarking work for AI systems is a strong plus.
• Contributions to open source projects are also a plus.
• Familiarity with agentic systems, large language model inference, or developer tooling is preferred.

• Competitive salary and performance bonuses.
• Flexible working hours and remote work options.
• Collaborative and innovative work environment with a flat organizational structure.
• Opportunities for professional development and continuous learning through workshops and conferences.
• Access to the latest tools and technologies in AI research.