About this role
Circana seeks a Principal Engineer - AI to lead the design and production delivery of agentic systems built around consumer-behavior data used by brands and retailers. The role works with domain experts to turn their knowledge into agents and reusable engineering patterns, while setting technical direction for a pod of engineers throughout a solution build.
Responsibilities
- Lead a solution family from domain discovery through production use. Translate subject-matter expertise into intent ontologies, tool contracts, and evaluation sets.
- Own agent harness engineering, including context management, MCP servers, guardrails, retries, graceful degradation, tracing, and replay.
- Make evaluation part of the development infrastructure through golden traces, regression suites, routing tests, and offline and online measurement. Establish reusable shared services, tools, and conventions.
- Review engineers’ work and communicate technical decisions to product owners, architects, security teams, executives, and client stakeholders.
Required qualifications
- A strong track record of designing and shipping agentic systems for real users, with production fluency in LangGraph, Google ADK, or Microsoft Agent Framework.
- Production experience with multiple model families and at least one major cloud AI stack; sound judgment in selecting models.
- Mastery of context management, prompt design, plan generation, structured outputs, and tool design. Daily fluency with agentic coding tools is required; Claude Code is preferred.
- Command of a development lifecycle using specifications as the source of truth, evaluations as continuous integration, prompt and tool versioning, and trace-driven debugging. Ability to communicate with engineers, domain experts, and senior executives.
Preferred background includes consumer goods, retail, or large-scale measurement data; forward-deployed or field engineering at an AI or data company; and public talks, open-source work, or technical writing. The listed stack includes Anthropic, Google, OpenAI, and open-weight model families; GCP Vertex AI and Azure AI Foundry; MCP, A2A, AG-UI, and A2UI; Python and TypeScript; and tracing and observability tools. The job listing identifies Bangalore and Pune, India, as locations.