Dataset Operations Manager

DoD Solutions Responds Quickly
$
🪖 DefTech Product

Location: On-site
Type: Full-time
Language Requirements: English (Ukrainian)

About the Role

High-quality ML systems depend on high-quality data, and producing that data is first and foremost about working with people. You will find and coordinate contributors, set clear expectations, review their work, and keep the process moving. You will run the data operations day to day: annotators, the drone operators who record video, and annotation vendors when we use them. The ML engineers decide what the data must contain and build the tools. You make sure the data arrives on time, with correct labels and full records.

Responsibilities

  • Turn ML team requests into clear annotation tasks, quality criteria, and deadlines. Agree priorities with the ML lead.
  • Run annotation projects in Label Studio: set up a project for each batch and export the finished annotations.
  • Find, onboard, and coordinate freelance annotators and, when we use one, an annotation vendor.
  • Plan workloads and budgets, assign work, track progress and costs, resolve blockers, manage rework, and ensure freelancers are paid accurately and on time. Collect recordings from drone pilots through the upload tool the engineers provide: train operators to use it, check file names, and follow up missing files.
  • When the engineers need new data, organize photo and video shoots, including contributors, locations, equipment, and logistics.
  • Coordinate synthetic dataset production with engineers and external contributors: plan delivery, source 3D assets, and check outputs against agreed requirements.
  • Create annotation guidelines and examples that produce consistent results. Bring unclear cases to the engineers and add their answers to the guidelines.
  • Review completed work, identify quality issues, and accept or return submissions. An ML engineer gives the final approval for each dataset.
  • Keep a record for each dataset: where it came from, who may use it, and which version is current. Release approved datasets with the team’s checklist, and make sure test recordings never end up in training data.
  • Keep the ML team informed and improve the process based on model results and changing priorities.

Required Qualifications

  • Experience coordinating people and detail-heavy operational work.
  • Experience coordinating projects with engineering teams.
  • Strong attention to detail and ability to review visual data consistently.
  • Ability to turn ambiguous requests into precise, practical instructions.
  • Confidence with spreadsheets, task-tracking tools, and annotation platforms.
  • Clear communication, reliable follow-through, and ownership of deadlines and results.
  • Hands-on experience delivering annotated photo or video datasets: coordinating annotators, writing instructions, and running quality checks; experience with Label Studio is a big plus.
  • DVC or basic Python is a plus. You do not need an ML background; a basic understanding of ML terms is enough. We will teach you our versioning and data checks.

What We Offer

  • The opportunity to build the data foundation for mission-critical defense technology
  • A fast-paced result-oriented culture where decisions are driven by data and results
  • Competitive compensation
  • Direct collaboration with the ML engineering team
  • Direct impact on Ukraine’s national security and technological edge
  • You report to the ML lead, who sets the priorities.

Required skills

Python

Required languages

English B2 - Upper Intermediate
Ukrainian Native
Published 16 November
10 views · 0 applications
To apply for this and other jobs on Djinni login or signup.
Loading...