Draup - Data Czar Engineer - Python/NLP

About Draup Draup is an AI-first market and talent intelligence platform trusted by 5 of the Fortune 10 and 300+ global enterprises including Microsoft, PepsiCo, PayPal, and Pfizer. Every day we analyze 1B+ data points from 100,000+ sources, powered by 150+ machine learning models and agentic workflows, to help enterprises bridge the gap between data and decisions. Our intelligence is only as good as the data beneath it and that is where you come in.

Why This Role Exists We are hiring a Data Czar : the single, accountable owner of how good Draups data is. Not a maintainer an owner. You will set the data agenda, hold the line on quality across our entire taxonomy, and push us toward better algorithms and models. And you will do it AI-first wielding prompting, agentic workflows, and evals to research and QA at a scale that was impossible a year ago. If you believe data is a product (not a byproduct) and you cannot walk past a flawed dataset without fixing it, this role was written for you.

What Youll Own

• The data research agenda. Decide what data Draup should capture, from which sources, and how it should be structured across 27K+ skills, 3,500+ roles, 6,100+ locations, and 33 industries. Turn open questions about the labor and market landscape into research that ships.
• Dataset quality, end to end. Be the last line of defense on accuracy, coverage, freshness, and integrity. Audit and QA existing datasets, hunt down duplicates, stale entities, and invalid records, and design the checks automated and human-in-the-loop that keep every entity (buyer, role, skill, company, technology) clean and trustworthy.
• Better algorithms and models. Continuously evaluate how we extract, classify, match, and infer. Benchmark what we run today, spot where it underperforms, and recommend stronger algorithms, ML models, and GenAI/LLM approaches then partner with Data Science and Engineering to put them into production.
• AI-accelerated research and QA. Put LLMs and agentic workflows to work on the grind. Design prompts, evals, and human-in-the-loop pipelines that let you research and quality-check at a scale no manual process can match and build the guardrails that keep AI honest (grounded outputs, no hallucinations, no drift).
• Data governance and trust. Champion bias mitigation, statistical rigor, and responsible data practices so our insights hold up to enterprise scrutiny and global standards such as SOC 2, GDPR, and ISO 27001.
• The standard. Define what good data means at Draup and make it measurable turning quality from a gut feel into metrics the whole team rallies around.

How Youll Work (AI-First) This is an AI-first role. We expect you to use AI to multiply yourself, not to replace your judgment.

• Prompting as a craft. You write precise, structured prompts and iterate like an engineer not a tourist who types a question and hopes.
• AI-assisted research. You use LLMs and agentic tools to gather, structure, and pressure-test information fast, while knowing exactly where they mislead you.
• Model evaluation and guardrails. You design evals and checks that catch hallucinations, drift, and bias before they ever reach a customer.
• Agentic workflows. You stitch tools, data, and models including MCP-style integrations into workflows that do real work end to end.
• Judgment over autopilot. You treat AI output as a draft to verify, never a source of truth. Human-in-the-loop is your default setting.

What Youll Bring

• A genuine love of data you find satisfaction in messy datasets made clean and hard questions made answerable.
• Strong hands-on experience in data research, data quality/QA, data science, or analytics engineering (typically 5-9 years).
• Fluency with Python and SQL and modern data tooling; comfort digging into large, real-world datasets.
• A working understanding of machine learning and GenAI/LLMs enough to evaluate models critically, reason about trade-offs, and recommend what is better and why.
• Hands-on fluency with AI tools (e.g., ChatGPT, Claude) and a real, day-to-day prompting practice you already use AI to do your job better.
• Sharp analytical judgment and an eye for the anomaly everyone else missed.
• The ability to operate with little direction and a lot of ownership.

Bonus Points

• Experience with NLP, entity resolution, taxonomy/ontology design, or large-scale ETL.
• Hands-on experience building RAG pipelines, LLM evals, embeddings/vector search, or fine-tuning.
• Familiarity with agentic frameworks or the Model Context Protocol (MCP).
• Background in labor market, talent, or B2B market intelligence data.
• A track record of building data-quality frameworks or QA processes from scratch. (ref:hirist.tech)