Data Operations Lead

Tractian; São Paulo; Híbrido; BRL 80.000 - 120.000; Tempo integral

Resumo da oferta

A leading data-driven company in São Paulo is looking for a Data Operations Lead to manage the lifecycle of data annotation programs. This hybrid role involves project management and quality assurance, requiring you to build the framework that supports AI development. Candidates must have 3+ years in Data Operations or related fields and proficiency in data tools like Excel. You will oversee a team of annotators and collaborate with external vendors to ensure high-quality data delivery.

Qualificações

• 3+ years in Data Operations, Program Management, or QA for Machine Learning/AI.
• Familiar with AI lifecycle.
• Experience writing technical documentation.
• Advanced proficiency in Excel/Google Sheets.
• Hands-on with annotation platforms.

Responsabilidades

• Define program objectives and deliverables for data projects.
• Create scalable SOPs and guidelines.
• Identify bottlenecks and implement risk strategies.
• Recruit and manage a team of annotators.
• Negotiate with external data vendors.
• Analyze errors and provide retraining materials.
• Maintain a master answer key for testing.
• Track and report on key quality metrics.
• Investigate issues affecting model performance.
• Help set up internal platforms for labeling.
• Generate executive dashboards for stakeholders.

Conhecimentos

Data Operations

Project Management

Quality Assurance

Data Analysis

Excel

SQL

Ferramentas

Labelbox

Scale AI

Appen Global

CVAT

Descrição da oferta de emprego

Data Science at TRACTIAN

The Data Science team at TRACTIAN focuses on extracting valuable insights from vast amounts of industrial data. Using advanced statistical methods, algorithms, and data visualization techniques, this team transforms raw data into actionable intelligence that drives decision-making across engineering, product development, and operational strategies. The team constantly works on optimizing prediction models, identifying trends, and providing data-driven solutions that directly enhance the company’s operational efficiency and the quality of its products.

What you’ll do

We are looking for a strategic Data Operations Lead to own the end-to-end lifecycle of our data annotation programs. You will not just manage tasks; you will build the “Ground Truth” engine that powers our AI. This role is a hybrid of Project Management and Quality Assurance, requiring you to orchestrate workflows between internal teams, external vendors, and automated tools. You will be responsible for defining the strategy, ensuring quality at scale, and delivering datasets that meet strict Service Level Agreements (SLAs).

Responsibilities

• End-to-End Ownership: Define program objectives, timelines, and deliverables for multiple data labeling projects simultaneously.
• Workflow Design: Create scalable SOPs (Standard Operating Procedures) and guidelines. You will decide when to use a “Consensus” model (multiple labelers per item) versus a “Single-Pass” model based on cost/quality trade-offs.
• Risk Mitigation: Proactively identify bottlenecks (e.g., ambiguity in guidelines, tool downtime) and implement “Risk Mitigation Strategies” before they impact model training schedules.
• Crowd/Team Oversight: Recruit, train, and manage a distributed team of annotators. Monitor “Throughput” (items/hour) and “Efficiency” to ensure productivity targets are met.
• Vendor Relations: Act as the primary interface for external data vendors. Negotiate timelines, track budget utilization, and hold vendors accountable to accuracy SLAs (e.g., 98% quality on Gold Sets).
• Performance Coaching: Implement data-driven feedback loops. If an annotator’s quality drops, you will analyze their errors and provide targeted retraining materials.
• Gold Set Management: Maintain a “Gold Set” (master answer key) to blindly test annotators.
• Metric Analysis: Track and report on key quality metrics: Inter-Annotator Agreement (IAA), Accuracy, and Precision/Recall of the human labels.
• Root Cause Analysis: When model performance dips, you will investigate the training data to determine if the issue stems from “Labeler Bias,” “Guideline Drift,” or “Edge Case Ambiguity.”
• Platform Operations: Help set up the UI/UX and configurations for the internal platforms used for labeling and annotation.
• Reporting: Generate weekly executive dashboards using Excel/Google Sheets (Pivot Tables, VLOOKUP) to visualize “Spend vs. Output” and “Quality Trends” for stakeholders.

Requirements

• Experience: 3+ years in Data Operations, Program Management, or QA for Machine Learning/AI.
• Technical Literacy: Familiarity with the AI lifecycle (Training vs. Validation vs. Test sets).
• Operational Rigor: Experience writing technical documentation/guidelines that leave no room for interpretation.
• Data Skills: Advanced proficiency in Excel/Google Sheets. (SQL experience is a strong plus).
• Tool Proficiency: Hands‑on experience with annotation platforms (e.g., Labelbox, Scale AI, Appen Global, CVAT).