Agentic AI/ML Engineer Intern, Solutions

FieldAI•Irvine, CA

1d•$35 - $50•Onsite

About The Position

FieldAI's Irvine team is focused on embodied AI, working with real robots, sensors, and field deployments. They build risk-aware, reliable, field-ready AI systems to solve complex robotics problems and unlock the potential of embodied intelligence. This role involves building agentic solutions that leverage FieldAI's data and tooling to provide actionable support and insights. The work has a dual focus: creating internal agents to accelerate engineering and operations workflows, and building external agentic experiences for customers to gain insights from their deployments. A key component is the AI Ops platform, which involves developing core infrastructure for orchestration, tool integration, memory, and observability. This is a hands-on role requiring prototyping, evaluation, and shipping agent-native solutions to multiply the impact of teams and technology. The role is crucial for scaling the company as the fleet, customer base, and team grow, enabling support for more customers and surfacing more insights without linearly scaling headcount. This is a paid internship with a strong potential for conversion to a full-time role.

Requirements

Currently pursuing a BS, MS, or Ph.D. in Computer Science, AI/ML, Robotics, or a related technical field, with deep project-based experience.
Strong evidence of building agentic projects (hackathons, research, internships, or personal projects).
Solid theoretical understanding and practical application of Agentic Engineering principles (Tool Use, Memory, RAG, Planning).
Proven ability to write reliable, testable, clean, and performant Python code, with familiarity with software engineering best practices, including version control, containerization (Docker), and test-driven development (pytest).
Hands-on engineering experience with modern, open-source agentic frameworks (e.g., LangChain, LangGraph, LlamaIndex) rather than relying strictly on service-managed agent APIs.
Experience implementing evaluation, tracing, and monitoring pipelines (e.g., MLflow, Langfuse, TruLens) to quantitatively measure agent quality, factual accuracy, latency, and reliability.
Practical expertise building and optimizing context-aware systems, with hands-on experience using Vector Databases (e.g., Pinecone, FAISS, OpenSearch) and designing Knowledge Graphs to reliably ground agents and mitigate hallucinations.
Familiarity with Cloud Platforms (e.g., AWS, GCP) for ML/AI deployment, and/or experience with on-robot compute environments.
Ability to take a loosely defined, complex problem and define and drive a working solution end-to-end.
Strong ability to drive solutions end-to-end, including cross-team coordination and seeking out customer input to shape what gets built.
Strong communication, initiative, and ability to learn quickly in a fast-moving team.

Nice To Haves

Deep experience designing and operating AI Ops infrastructure at production scale, including robotics-grade data logging and observability (e.g., Foxglove).
Experience with advanced agent patterns: Multi-Agent Systems, Human-in-the-Loop workflows, or Long-Horizon Planning.
Prior experience shipping internal tools or customer-facing assistants used by real users.
Personal projects and a portfolio of agentic builds are a big bonus — we love seeing what you’ve shipped on your own.

Responsibilities

Design and implement agentic workflows with tool use, memory, and orchestration to automate repetitive tasks and answer questions over internal and customer-facing data.
Contribute to AI Ops (agent infrastructure) — orchestration, evals, and observability — and apply it to enable agent-native DevOps that automates our engineering and internal operations workflows.
Build and optimize RAG pipelines with vector DBs and knowledge graphs to ground agents in the right context.
Set up evaluation pipelines to measure agent quality, reliability, and performance.