AI/ Machine Learning Engineer

1 month ago

Pune, IndiaMid Level / Senior

Responsibilities

Design and implement state-of-the-art ML and LLM-powered features for the Aera Platform.
Build and own end-to-end agentic workflows including multi-step reasoning, tool use, subagent orchestration, autonomous loops, and human-in-the-loop checkpoints.
Develop offline and online evaluation harnesses, golden datasets, regression gates, and non-deterministic test suites.
Optimize agent performance and cost through context engineering, prompt caching, asynchronous tool calls, model routing, structured outputs, and token budgeting.
Build tracing and observability systems to profile agent behavior, identify bottlenecks, and detect regressions.
Design model-usable tool and skill surfaces, including MCP servers and reusable agent skills.
Integrate models into robust data pipelines, inference paths, and serverless infrastructure for enterprise workloads.
Implement agent security controls including tool permissioning, sandboxing, and prompt-injection resistance.
Collaborate with Data Science, Engineering, and DevOps, and evaluate emerging AI techniques.

B.E. or B.Tech in Computer Science, Computer Engineering, or a related field.
3–5 years of software engineering and architecture experience.
At least 2 years designing and deploying ML or LLM-based systems, including 6–12 months of LLM-specific work.
Experience architecting and owning complex, high-stakes systems from design through production reliability.
Strong systems thinking and experience abstracting cross-domain problems into reusable platform building blocks.
Strong Python and production-service experience with FastAPI or an equivalent framework.
Experience with large datasets, ML pipelines, and distributed systems such as Ray or Spark.
Hands-on experience with PyTorch, Hugging Face, scikit-learn, and pandas.
Experience with containerized services using Docker and Kubernetes.
Experience with Git, Jenkins, and Jira.
Fluency with agentic engineering practices including context engineering, tool and skill design, subagent patterns, agent memory, evaluations, LLM-as-judge, structured outputs, prompt caching, and RAG.
Experience using agentic coding tools such as Claude Code or equivalent.
Preferred: GoLang, vector databases, event streaming, caching, Temporal, agent observability and experiment tracking tools, fine-tuning, multimodal AI, and AWS, GCP, or Azure serverless AI infrastructure.

Competitive salary and company stock options are offered, but no salary amount is stated.
Medical, Group Medical Insurance, Term Insurance, Accidental Insurance, paid time off, and maternity leave are available.
Unlimited access to online professional courses and people-manager development programs is provided.
The company offers a flexible working environment and a fully stocked office kitchen with snacks and beverages.
The role is based in the Pune office.

201-500 employees