Job description

FieldAI ’s Irvine team is where embodied AI meets real robots, real sensors, and real field deployments. Based in the heart of Southern California’s robotics ecosystem, we build risk-aware, reliable, field-ready AI systems that solve the hardest problems in robotics and unlock the full potential of embodied intelligence. If you want your work to ship, get tested on hardware, and improve through real deployments, Irvine is the place. We go beyond typical data-driven approaches or pure transformer-only architectures, combining rigorous engineering with learning systems proven in globally deployed solutions that deliver results today and get better every time our robots run in the field. About the Job Our Field Foundation Model (FFM) powers a global fleet of autonomous robots that capture massive streams of multimodal data across diverse, dynamic environments every day. As part of the Insight Team our mission is to transform this raw, multimodal data into actionable insights that empower our customers and engineers to deliver value. Field-insight Foundation Model (FiFM) is at the core of how we transform multimodal data from autonomous robots into actionable insights. As an AI/ML Engineer on the FiFM team, you will drive research and model development for one of Field AI’s most ambitious initiatives. Your work will span computer vision, vision-language models (VLMs), multimodal scene understanding, and long-memory video analysis and search, with a strong emphasis on agentic AI (tool use, memory, multimodal retrieval-augmented generation).This is a full-cycle ML role: you’ll curate datasets, fine-tune and evaluate models, optimize inference, and deploy them into production. It’s a blend of applied research and engineering, requiring creativity, rapid experimentation, and rigorous problem-solving. While FiFM is your primary focus, you’ll also contribute to broader perception and insight-generation initiatives across Field AI. What You’ll Get To Do Train and fine-tune million- to billion-parameter multimodal models, with a focus on computer vision, video understanding, and vision-language integration. Track state-of-the-art research, adapt novel algorithms, and integrate them into FiFM. Curate datasets and develop tools to improve model interpretability. Build scalable evaluation pipelines for vision and multimodal models. Contribute to model observability, drift detection, and error classification. Fine-tune and optimize open-source VLMs and multimodal embedding models for efficiency and robustness. Build and optimize Multi-VectorRAG pipelines with vector DBs and knowledge graphs. Create embedding-based memory and retrieval chains with token-efficient chunking strategies. What You Have Master’s/Ph.D. in Computer Science, AI/ML, Robotics, or equivalent industry experience. 2+ years of industry experience or relevant publications in CV/ML/AI. Strong expertise in computer vision, video understanding, temporal modeling, and VLMs. Proficiency in Python and PyTorch with production-level coding skills. Experience building pipelines for large-scale video/image datasets. Familiarity with AWS or other cloud platforms for ML training and deployment. Understanding of MLOps best practices (CI/CD, experiment tracking). Hands-on experience fine-tuning open-source multimodal models using HuggingFace, DeepSpeed, vLLM, FSDP, LoRA/QLoRA. Knowledge of precision tradeoffs (FP16, bfloat16, quantization) and multi-GPU optimization. Ability to design scalable evaluation pipelines for vision/VLMs and agent performance. The Extras That Set You Apart Experience with Agentic/RAG pipelines and knowledge graphs (LangChain, LangGraph, LlamaIndex, OpenSearch, FAISS, Pinecone). Familiarity with agent operations logging and evaluation frameworks. Background in optimization: token cost reduction, chunking strategies, reranking, and retrieval latency tuning. Experience deploying models under quantized (int4/int8) and distributed multi-GPU inference. Exposure to open-vocabulary detection, zero/few-shot learning, multimodal RAG. Knowledge of temporal-spatial modeling (event/scene graphs). Experience deploying AI in edge or resource-constrained environments. Our salary range is generous and we consider each individual’s background and experience when determining final compensation. Base pay may vary based on role scope, job-related knowledge, skills, experience, and the Irvine, California market. Why Join FieldAI in Irvine? In Irvine, you will work where the robots are. Our local team builds and tests systems on real hardware with real sensors, then ships them to operate in unstructured, previously unknown environments around the world. We are solving one of robotics’ hardest challenges: reliable deployment outside the lab. Our Field Foundational Models™ raise the bar for perception, planning, localization, and manipulation, with an emphasis on explainability and safety for real-world use. You will collaborate with a world-class team that thrives on creativity, resilience, and bold thinking. We bring deep experience from organizations such as DeepMind, NASA JPL, Boston Dynamics, NVIDIA, Amazon, Tesla Autopilot, Cruise, Zoox, Toyota Research Institute, and SpaceX, along with a track record of field deployments and strong performance in DARPA challenge segments. Be Part of the Next Robotics Revolution We are looking for builders who want their work to leave the whiteboard and show up on robots. If you enjoy tackling tough, uncharted questions and working across disciplines, you will find your people here. Our teams span AI, software, robotics engineering, product, field deployment, and technical communication, all focused on shipping systems that perform in the real world. Our headquarters is in Irvine, and we partner closely with teams there as well as colleagues across the US and around the world. Join us in Southern California and help define what dependable, field-ready autonomy looks like. We value diverse perspectives and are committed to fostering an inclusive workplace. We evaluate candidates and employees based on merit, qualifications, and performance, and we do not discriminate on the basis of race, color, gender, national origin, ethnicity, veteran status, disability status, age, sexual orientation, gender identity, marital status, or any other legally protected status.

Requirements

Must-haves

What You’ll Get To Do
Train and fine-tune million- to billion-parameter multimodal models, with a focus on computer vision, video understanding, and vision-language integration.
Track state-of-the-art research, adapt novel algorithms, and integrate them into FiFM.
Curate datasets and develop tools to improve model interpretability.
Build scalable evaluation pipelines for vision and multimodal models.
Contribute to model observability, drift detection, and error classification.
Fine-tune and optimize open-source VLMs and multimodal embedding models for efficiency and robustness.
Build and optimize Multi-VectorRAG pipelines with vector DBs and knowledge graphs.
Create embedding-based memory and retrieval chains with token-efficient chunking strategies.
What You Have
Master’s/Ph.D. in Computer Science, AI/ML, Robotics, or equivalent industry experience.
2+ years of industry experience or relevant publications in CV/ML/AI.
Strong expertise in computer vision, video understanding, temporal modeling, and VLMs.
Proficiency in Python and PyTorch with production-level coding skills.
Experience building pipelines for large-scale video/image datasets.
Familiarity with AWS or other cloud platforms for ML training and deployment.
Understanding of MLOps best practices (CI/CD, experiment tracking).
Hands-on experience fine-tuning open-source multimodal models using HuggingFace, DeepSpeed, vLLM, FSDP, LoRA/QLoRA.
Knowledge of precision tradeoffs (FP16, bfloat16, quantization) and multi-GPU optimization.
Ability to design scalable evaluation pipelines for vision/VLMs and agent performance.
The Extras That Set You Apart
Experience with Agentic/RAG pipelines and knowledge graphs (LangChain, LangGraph, LlamaIndex, OpenSearch, FAISS, Pinecone).
Familiarity with agent operations logging and evaluation frameworks.
Background in optimization: token cost reduction, chunking strategies, reranking, and retrieval latency tuning.

Nice-to-haves

None flagged.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

aigomlpythonawsci/cd

Platform: lever
Effort: quick

Benefits

Benefits detected in the listing.

Vision