About the job
ML/AI Research Engineer - Founding Team at Agentic AI Lab
Location: San Francisco Bay Area
Type: Full-Time
Compensation: Competitive salary + meaningful equity (founding tier)
At fabrion, backed by 8VC, we are assembling a top-tier team dedicated to addressing one of the most pressing infrastructure challenges in the industry.
About the Role
Join us in shaping the future of enterprise AI infrastructure, focusing on agents, retrieval-augmented generation (RAG), knowledge graphs, and multi-tenant governance.
As an ML/AI Research Engineer, you will spearhead the design, training, evaluation, and optimization of agent-native AI models. Your work will integrate LLMs, vector search, graph reasoning, and reinforcement learning, establishing the intelligence layer for our enterprise data fabric.
This role goes beyond prompt engineering; it encompasses the entire ML lifecycle, from data curation and fine-tuning to thorough evaluation, interpretability, and deployment, all while considering cost-effectiveness, alignment, and agent coordination.
Core Responsibilities
Fine-tune and assess open-source LLMs (e.g., LLaMA 3, Mistral, Falcon, Mixtral) for enterprise applications, leveraging both structured and unstructured data.
Construct and enhance RAG pipelines utilizing LangChain, LangGraph, LlamaIndex, or Dust, integrating with our vector databases and internal knowledge graphs.
Train agent architectures (ReAct, AutoGPT, BabyAGI, OpenAgents) using enterprise task datasets.
Develop embedding-based memory and retrieval chains employing token-efficient chunking strategies.
Create reinforcement learning pipelines to enhance agent behaviors (e.g., RLHF, DPO, PPO).
Establish scalable evaluation harnesses for LLM and agent performance, including synthetic evaluations, trace capture, and explainability tools.
Contribute to model observability, drift detection, error classification, and alignment efforts.
Optimize inference latency and GPU resource utilization across both cloud and on-premises environments.
Desired Experience
Model Training:
Deep understanding of machine learning principles and hands-on experience with model training.
