Principal AI Engineer - Context - Agents and Context
About this role
The Context Engine team at Elastic builds the knowledge layer that AI agents use to interact with enterprise data stored within Elasticsearch. This engineer owns the end-to-end product improvement loop, focusing on how agents, automations, and skills behave in production environments, alongside designing evaluation systems and telemetry. The position operates at the intersection of engineering, data science, and product development, requiring someone who can write production code in TypeScript, establish reliability standards, and collaborate closely with cross-functional peers in a distributed team.
Candidates joining this team will shape how knowledge is extracted from diverse sources into structured AI indices and served through public APIs, MCP tools, and framework integrations. Ideal candidates bring extensive engineering backgrounds with a strong focus on eval-driven product improvement, handling agent state and memory, and maintaining external contracts as data models evolve. A pragmatic and low-ego approach is essential for thriving in this async-first environment.
What you will do
- Manage the production improvement loop for the Context Engine by analyzing extraction automations, retrieval tools, and memory through offline evaluations and telemetry.
- Establish safety protocols for agent and skill iteration, including versioning, rollout strategies, regression coverage, and staged evaluations.
- Design telemetry systems to capture data from agent traces, tool calls, and knowledge retrieval to support product decisions.
- Collaborate with the data science team to define evaluation strategies such as golden datasets and quality evaluators.
- Guide the team by reviewing designs and pull requests, mentoring engineers, and authoring technical proposals.
- Write production code in TypeScript and contribute to the Elastic Stack codebase in a public, open development model.
What you need
- Possess more than 10 years of software engineering experience with recent focus on shipping and operating AI-driven products on production traffic.
- Demonstrate a track record of eval-driven product improvement by diagnosing agent or LLM behavior from traces and user feedback.
- Bring direct experience building agents with state and memory while iterating on prompts, skills, and tool behavior safely in production.
- Show familiarity with the Model Context Protocol (MCP), including the exposure of public MCP servers and tools.
- Bring experience in designing telemetry for AI systems and running end-to-end product experiments.
- Demonstrate strong backend engineering skills in either Python or TypeScript for building APIs, stateful workflows, and data pipelines.
Nice to have
- Experience with agent frameworks such as LangGraph, CrewAI, Claude Agent SDK, or similar
- Agentic retrieval experience, including knowledge representation
- Practical Elasticsearch experience
Benefits
- Competitive pay based on current work
- Health coverage for employee and family in many locations
- Flexible locations and schedules for many roles
- Generous number of vacation days each year
- Matching up to $2000 for financial donations and service
- Up to 40 hours each year for volunteer projects and a minimum of 16 weeks of parental leave
Skills and tech stack
Questions candidates ask
Is this a remote position?
This is a hybrid role with flexible locations and schedules available for many roles within a distributed, async-first team.
What programming languages are used in the codebase?
The codebase is primarily TypeScript, with strong backend engineering skills required in either Python or TypeScript.
What benefits are provided for this role?
Benefits include competitive pay, health coverage for employees and family in many locations, flexible schedules, a generous number of vacation days, donation matching up to $2000, volunteer time off, and a minimum of 16 weeks of parental leave.