- Location
- Santa Clara, California
- Workplace
- Remote
- Type
- Full-time
- Department
- IT
- Seniority
- Senior
- Source
- Ashby
Description
Who we are
Odyssey is an AI lab pioneering general-purpose world models: causal, multimodal systems that learn to predict and interact with the world over long horizons. This foundational technology promises to revolutionize robotics, science, healthcare, education, gaming, defense, and beyond.
Odyssey is an AI lab pioneering general-purpose world models: causal, multimodal systems that learn to predict and interact with the world over long horizons. This foundational technology promises to revolutionize robotics, science, healthcare, education, gaming, defense, and beyond.
What we’re looking for
The Data Program Manager is the connective tissue between research, vendors and the data flywheel. You are responsible for understanding our data needs by being deep in the frontier research space for multi-model methodologies and arxiv papers. You are responsible for presenting a credible data roadmap to the research, applied ML, and leadership and keeping stakeholders abreast of the pipeline and plans. You also own inbound requirements and turn those into a backlog the data-ops team can execute against.
The right person thinks in process, has strong opinions, and can hold both people (internal and external) and plans accountable.
What you’ll do
Run at data plans with loose requirements from research. Turn fuzzy and disparate data requests into scoped, sequenced acquisition plans with quality bars, timelines, and clear data parameters
Manage a multitude of vendors across verticals and data types, maintain relationships and negotiate commercials favorable to Odyssey
Manage the human teams (experience with contractors a big plus) that acquire proprietary data and establish and ensure quality measures
Stand up and own the weekly data-needs review and forecast. Force-prioritize when necessary and surface the trade-offs to leadership.
Embed in research planning. Flag data risks four to six weeks ahead of when they would hit a model run.
Define and track quality and throughput metrics across active data programs. Redline and QA batches that slip and close the loop with vendors.
Partner with Research on data requirements while also running independently at frontier data mix delivery
Drive program retrospectives – what data moved which evals, at what cost, and what we should repeat.
Who you are
5+ years of program or product management at a technical company, with at least two of those years in an AI, ML, or data-intensive context.
You can read ML papers without flinching; you don’t need to derive the math, but you understand what the research is trying to prove and what data would support it
Working knowledge of SQL and basic data analysis. You can pull your own numbers when needed
Real experience negotiating with external vendors and pushing costs down
You can run a tight standup, keep clean docs, and finish things
Comfort saying “no” to a researcher; better at saying “yes, by week six, here’s the plan”
Track record of driving cross-functional programs at velocity in environments where priorities change weekly
Bonus
Experience as a TPM or program manager at a frontier AI lab
Background working with data labeling, human-data collection, or RLHF programs at scale
Familiarity with AI evaluation methodology or rubric design
Prior experience as the first X hire at an early-stage company
