Technical Program Manager
[Evaratus] | in-office | Full-time
About the Role
Evaratus builds the operating infrastructure that turns large-scale, sensitive data into safe,
high-quality inputs for frontier AI labs. Every engagement we take on runs through a multi-stage
delivery pipeline that has to hold up reliably, on schedule, and at scale - often with several
workstreams moving in parallel.
We're hiring a Technical Program Manager to bring structure, visibility, and reliability to how we
run these programs end-to-end from a signed commitment through execution, quality
assurance, and delivery. This is a build-the-plane-while-flying-it role at an early-stage company:
you'll be establishing the operating rhythm as much as running it, and the skills you bring will
extend across whichever programs we take on next. Our current flagship program, an enterprise
data delivery pipeline covering extraction, PII redaction, and quality assessment is the first
proving ground for that operating model, not the limit of the role.
What You'll Own
Program & Delivery Management
- Drive complex, multi-stage programs through their full lifecycle - kickoff, execution,
- quality checks, and handoff — against committed timelines and SLAs.
- Own the cross-functional operating cadence: run recurring syncs across the teams involved, and keep decisions and owners logged, not just discussed.
- Design and run a formal risk/blocker escalation process with clear SLAs, so unresolved issues surface automatically instead of being chased down manually.
- Track team capacity against the active program pipeline and flag overload risk before it hits delivery dates.
- Identify and execute initiatives to improve processes and continuously build efficiency,
Drive Technical Prowess and Lean Machine
- Help design and operate QA loops for critical pipelines: structured human-in-the-loop sampling of live output, not just static test-set metrics, tracked against clear precision/recall or accuracy measures.
- Contribute to a defensible ground-truth process, a multi-annotator labeling, inter-annotator agreement scoring, and adjudication for disagreements, so "high confidence" is actually measured.
- Maintain an error taxonomy that turns one-off failures into real signals for the team, not just rework and Help define and enforce release gates for pipeline versions before they ship to live delivery.
- Build product, automation and efficiency iteratively to make high quality and reliability as a default standard of any delivery.
What We're Looking For
- 5+ years in technical program management, delivery/engineering program management, or a closely adjacent operations role ideally in a high-growth environment.
- A track record running multi-stage, cross-functional delivery pipelines against hard external SLAs, not just internal engineering roadmaps.
- Comfort building process and tooling from scratch (status rollups, escalation frameworks, capacity tracking) rather than inheriting mature systems.
- Excellent written and verbal communication, you'll be the translation layer between engineering, quality/operations teams, and account-facing stakeholders.
- Working fluency with data pipeline concepts: throughput, batching, versioning, and core data quality dimensions (completeness, consistency, duplication rate).
- High degree of ownership to ensure that delays are owned, deliveries are tracked, red-lines escalated, ambiguities disambiguated and reasonable calls made
How We Work
We value clear ownership. We're an early-stage team building foundational infrastructure for how the AI industry sources and trusts enterprise-scale data — the processes you put in place now are the ones the company scales on for years, across whatever programs come next. We welcome applicants from all backgrounds and career paths who bring rigor and ownership to ambiguous, high-stakes delivery problems.
