Principal AI Engineer - Context - Agents and Context

Elastic
Spain, United Kingdom
Workplace: HybridFull timeEUR 80,400 - 127,200Function: Data Science & Machine LearningExperience: 10+ yearsSkills: ["Low-ego collaboration","Mentoring","Designing evaluations","Data-informed decision-making","Communication"]

Own the end-to-end improvement loop for the Context Engine knowledge layer that powers AI agents in Elasticsearch. Build production code for extraction automations, retrieval tools, memory, and agent traces; design offline evaluations and telemetry; and safely ship prompt/skill/automation versioning with guardrails. Partner with data science, backend engineering, product, and UX to identify failure modes, measure outcomes, and raise team quality through eval-driven development.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Elastic
Elastic
22 hours ago

Principal AI Engineer - Context - Agents and Context

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 8 hours agoStatus: Live

Job Summary

Own the end-to-end improvement loop for the Context Engine knowledge layer that powers AI agents in Elasticsearch. Build production code for extraction automations, retrieval tools, memory, and agent traces; design offline evaluations and telemetry; and safely ship prompt/skill/automation versioning with guardrails. Partner with data science, backend engineering, product, and UX to identify failure modes, measure outcomes, and raise team quality through eval-driven development.
Location: Spain, United Kingdom
Workplace: Hybrid
Employment Type: Full time
Job Function: Data Science & Machine Learning
Seniority: Mid level

Key Responsibilities

  • •Own the production improvement loop for Context Engine, using offline evaluations and telemetry to identify failure modes, fix them, and prove the results.
  • •Define safe iteration processes for agents and skills, including versioning, rollout, regression coverage, and staged/shadow evaluation with guardrails.
  • •Design telemetry for agent traces, tool calls, and knowledge retrieval, and route it into Elasticsearch for evaluation and dashboards.
  • •Partner with data science to develop evaluation strategy, including golden datasets and evaluators for quality, latency, and cost.
  • •Raise team standards through design/PR reviews, mentoring engineers in eval-driven development, and writing technical proposals shaping the roadmap.

Pay and Benefits

Salary: EUR 80,400 - 127,200
Perks:Health InsurancePaid LeaveParental Leave

Key Requirements

  • •10+ years of software engineering experience, including shipping and operating AI-driven products on real production traffic.
  • •Experience with eval-driven product improvement: diagnose issues from traces/user feedback, design evaluations, ship fixes, and measure impact.
  • •Direct experience building agents with state and memory, and iterating on prompts, skills, and tool behavior safely in production.
  • •Familiarity with MCP, including exposing public MCP servers and tools.
  • •Strong backend skills in either Python or TypeScript, including APIs, stateful workflows, data pipelines, and production services.
Experience:10+ yearsAI-driven productsAgentsLLMEnterprise software
Skills:Low-ego collaborationMentoringDesigning evaluationsData-informed decision-makingCommunication
Languages:English
Tech Stack:TypeScriptPythonElasticsearchMCPLangChainClaude CodeLangGraphCrewAIClaude Agent SDK

Company Brief

Elastic
Builds the Elastic Stack (Elasticsearch, Kibana, Beats, Logstash) and provides search, observability, and security solutions that enable organizations to search, analyze, and protect data in real time across applications, infrastructure, and enterprises.
Industry: Data Infrastructure
Company Size: Enterprise (1,001+ employees)
Revenue: USD 1B+
Growth: Public Company
Valuation: Public Company (Market Cap in USD)
Funding: IPO / Publicly Listed
Headquarters: Amsterdam, Netherlands
Founded: 2012
WebsiteLinkedIn