Causaly Senior Agent Harness Engineer at Causaly designs and implements AI agents end-to-end, ensuring reliability and maintainability. Collaborates with cross-functional teams to ship AI-powered user-facing features.
Responsibilities
Design and implement AI agents end-to-end, from the first idea to deployment and monitoring, balancing cutting-edge techniques with pragmatism to deliver measurable impact.
Apply strong software engineering principles, such as modularity, testing, code reviews, CI/CD and observability, to ensure AI systems are reliable, maintainable, production-ready and can be readily adapted to future developments.
Choose the right approach for the problem at hand, evaluating different models, prompts, tools and agent designs to balance trade-offs between speed, cost, complexity, interpretability, and performance.
Collaborate closely with product, design, and other engineering teams to scope work, align on success metrics, and incrementally ship improvements in user-facing features powered by AI.
Document system architectures and decision rationale early and clearly, enabling alignment across teams and accelerating onboarding and iteration.
Champion model and data quality, including robust evaluation, hallucination checks, and real-world performance tracking.
Mentor junior AI engineers and cross-functional teammates, sharing best practices in modelling, coding, maintaining and integrating product features, and helping grow a high-trust, high-performance team culture.
Stay up-to-date with emerging research and tools, distilling key insights and bringing back relevant innovations to elevate team capabilities and product opportunities.
Contribute to a culture of knowledge sharing, through company-wide Slack channels, Show and Tell presentations and technical deep-dives.
What experience you’ll need to be successful
A master's degree or above in Computer Science, Electrical Engineering or a related field.
5+ years of experience building software in production environments, including LLM-based systems, with ownership of key lifecycle stages: design, evaluation, deployment, and monitoring.
Qualification
Solid understanding of LLMs and agents
Required
Master's degree or above in Computer Science, Electrical Engineering or related field
5+ years of experience building software in production environments including LLM-based systems
Proficiency in Python and modern agentic frameworks such as LangGraph or LangChain