Tech Lead Manager, AI Data Products

Mercor•San Francisco, CA

2d•$300,000 - $420,000•Onsite

About The Position

Mercor’s Frontier Data Products team builds systems that turn multimodal inputs—including images, video, audio, documents, and mixed media—into reliable, customer-ready data products. As Tech Lead Manager, you will lead the team responsible for delivering these products at scale. This is a player-coach role split roughly evenly between technical contribution and people leadership. You will write and review production code, own architecture decisions, resolve difficult production problems, and hire, coach, and organize a team with clear ownership and strong independent judgment. This is a product-engineering leadership role: success is measured by the quality, reliability, and usefulness of delivered products, not research output alone. The role is based in person five days a week in San Francisco, New York, or London. You will shape multimodal product architecture and the team building it, ensuring systems preserve context, relationships, provenance, and quality across modalities and transformations. The architecture should support new products and evolving customer needs without requiring a separate system for every use case.

Requirements

Bachelor’s degree in Computer Science or a related technical field, such as Electrical Engineering, Physics, Mathematics, or Statistics.
At least 1 year of experience managing a strong engineering team while remaining an effective, hands-on technical leader.
At least 3 years of deep experience designing and operating production backend, distributed, or data-intensive systems.
Experience working with large unstructured data, asynchronous processing, metadata and versioning, or compute-heavy workflows.

Nice To Haves

Relevant experience in multimedia infrastructure, computer-vision data, video or audio processing, ML platforms, annotation systems, or human-in-the-loop products.
Strong judgment about platform boundaries, including what should be generalized across modalities and what should remain modality-specific.
A track record of taking ambiguous, zero-to-one products through architecture, launch, and sustained production operation.
Evidence of recruiting, coaching, and retaining excellent engineers while maintaining a high performance bar.

Responsibilities

Set the technical and product strategy for Mercor’s multimodal data products.
Design systems for the full lifecycle of image, video, audio, document, and mixed-media data, from initial inputs through processing, review, validation, versioning, and delivery.
Create reusable abstractions across modalities while preserving modality-specific requirements for quality, performance, and customers.
Build quality systems combining automated evaluation, model-assisted checks, expert review, sampling, and adjudication.
Ensure outputs are traceable, including what produced them, what changed, how they were evaluated, and why they were accepted.
Own the reliability, performance, and cost of storage-heavy, compute-intensive, long-running workflows.
Write production code, lead design reviews, and directly contribute to the team’s most consequential technical work.
Manage and develop engineers through clear expectations, frequent feedback, thoughtful delegation, and meaningful ownership.
Partner with product, ML, operations, and customer-facing teams to turn customer needs into durable product capabilities.
Recruit and onboard engineers who raise the team’s technical and execution bar.