About the role
Responsibilities:
- Design and build production-grade LLM and agentic AI systems.
- Extensive experience in context engineering and harness engineering.
- Design agent workflows involving tools, APIs, retrieval, memory, policies, and human-in-the-loop escalation.
- Build context engineering layers that retrieve, rank, structure, and manage business context for LLM applications.
- Good to have exposure to the Rust programming language.
- Develop prompt, tool, and workflow orchestration patterns for reliable AI behaviour.
- Create evaluation harnesses for LLM applications, including regression tests, golden datasets, automated scoring, trace replay, and human review workflows.
- Define and track metrics for task success, accuracy, cost, latency, safety, and reliability.
- Integrate LLM systems with enterprise data sources, APIs, vector databases, search systems, and operational platforms.
- Implement observability for AI systems, including tracing, prompt/version tracking, tool-call monitoring, and failure analysis.
- Partner with product, data science, engineering, analytics, security, and business teams to convert ambiguous problems into scalable AI solutions.
- Lead architecture reviews, mentor engineers, and establish best practices for LLM application development.
- Evaluate models, frameworks, vendors, and infrastructure options based on quality, latency, cost, privacy, and maintainability.
Requirements:
- 7 - 9 years of experience in Machine learning and data modelling.
- Strong experience in machine learning engineering, software engineering, data science engineering, or AI platform development.
- Hands-on experience building production ML, LLM, NLP, search, recommendation, or decision-support systems.
- Strong Python programming skills and experience building production services or APIs.
- Experience with LLM application patterns such as RAG, tool calling, structured outputs, workflow orchestration, or AI agents.
- Experience designing evaluation frameworks for ML or LLM systems.
- Understanding of prompt engineering, context design, retrieval quality, and model behaviour testing.
- Experience with cloud infrastructure, CI/CD, monitoring, logging, and deployment practices.
- Ability to reason about trade-offs across accuracy, latency, cost, reliability, privacy, and user experience.
- Strong communication skills and ability to lead technical work across teams.