Technical Program Manager, Training Data

Brain Co.•San Francisco, CA

About The Position

Brain Co. is looking for a TPM to own the end-to-end labeling process across our AI/ML initiatives. This person will serve as the single point of ownership for labeling operations, partnering closely with ML engineers, labelers, vendors, and subject matter experts to ensure high-quality labeled data is delivered efficiently, on time, and at scale. This is a highly cross-functional role for someone who combines strong operational ownership with technical fluency. You will design and improve labeling workflows, manage the people and processes behind them, and continuously raise the bar on quality, speed, and scalability.

Requirements

2+ years of experience in dataset curation, data labeling operations, data annotation, or a similar role
Strong ownership mindset and comfort serving as the point person for a complex workflow
Excellent written communication skills, with the ability to create precise and usable instructions
Strong organizational and project management skills, with the ability to plan around timelines and dependencies
Proactive communicator who raises issues early and helps unblock progress
Sufficient technical knowledge to work effectively with ML and engineering teams
Experience interviewing, managing, or coordinating labelers, vendors, or subject matter experts across different domains
Comfort with lightweight scripting, prototyping, or automation to streamline workflows

Nice To Haves

Experience working with ML, data, or human-in-the-loop workflows
Familiarity with labeling platforms, annotation tooling, and QA processes
Experience in complex or regulated domains such as health, insurance, or permitting

Responsibilities

Own labeling operations end-to-end across multiple AI/ML projects
Source, interview, onboard, and manage labelers, agencies, and subject matter experts
Write, refine, and maintain clear labeling instructions, examples, and task definitions
Partner with engineers to determine the right labeling approach, including inputs, outputs, workflows, and quality requirements
Identify and organize the right data and example sets for labeling
Keep labeling work on track against project milestones and proactively communicate risks, blockers, and tradeoffs
Monitor quality, accuracy, and efficiency, and drive corrective actions when gaps appear
Recommend the right tools for data collection and labeling, and help shape lightweight in-house tooling or automation where needed
Facilitate collaboration between engineering and labeling teams to ensure alignment on scope, quality, and timelines
Own dataset curation workflows end-to-end, including working with production data, public data sources, and external SMEs
Build and improve long-term labeling processes with a focus on efficiency, latency, accuracy, and scalability