Tech Lead, Machine Learning: Element Classification

**About Teleskope** Teleskope is redefining data security for the AI era with the only dedicated platform that combines precise visibility with automated remediation. Teleskope continuously scans, catalogs and classifies data in-motion and at-rest while automating policy-based actions, helping organizations proactively manage data sprawl while securely enabling AI adoption. Following our $25 million Series A round, Teleskope is entering a high-growth phase backed by top-tier investors and exceptional product-market fit. **About the role** We're hiring a **Tech Lead** to own the technical direction of our element classification system, the engine that powers every customer workflow on the platform. You'll lead the engineers and data scientists building entity classification across text, documents, relational data, and OCR. You'll set the roadmap, decide where the pipeline needs to improve, and make sure our methods stay at the state of the art. This is a hands-on leadership role. You'll split your time between setting direction for the team and doing the hardest technical work yourself. You'll report to the Director of Data Science and ML and work closely with annotation, QC, engineering, and product. This is a hybrid **role requiring 3+ days in-office in New York City**. **What You'll Do**: - Own the element classification roadmap: set priorities, sequence the work, and align the team and stakeholders on tradeoffs. - Lead the team's technical work: break large problems into scoped projects, assign and unblock work, review designs and code, and hold a high bar for quality. - Own evaluation strategy: build benchmarks, regression suites, and production monitoring that show where the pipeline is failing and why, so the team focuses on the gaps that matter. - Track new work in NLP, LLMs, and information extraction, and decide what to adopt. Run focused experiments and take the winners to production. - Set the standard for data quality: refine labeling guidelines, work with annotators and qc teams on label quality, and build training and eval sets that reflect real world problems. - Run delivery with checkpointed milestones, and keep leadership and cross-functional partners updated on progress, risks, and decisions. - Set engineering practices for model deployment, versioning, CI/CD, and operational monitoring. - Mentor and grow the team, and help with hiring as the team scales. **About you:** - 6+ years building and deploying ML systems in production, including 1+ year as a tech lead or in a similar role guiding the technical work of others. - Deep experience with NLP, especially NER, entity classification, or information extraction. - Strong instincts for evaluation: you can build test sets that predict real-world performance and diagnose failures from metrics. - Hands-on experience curating datasets, including annotation guidelines, label quality control, and synthetic or weakly supervised data. - Up to date on current methods (LLMs, distillation, efficient inference) and able to judge them critically. - Still writes production code: comfortable taking work from research prototype to CI/CD and production pipelines in Python with modern ML tooling. - Turns ambiguity into a plan the team can execute, and communicates progress honestly to technical and non-technical stakeholders. - Leads through influence and makes the people around you better. **What you'll get:** - Technical ownership of the system behind the core of our product. - A team to lead and a direct hand in shaping how ML is built at Teleskope. - A beautiful, well-stocked office in NYC’s Financial District. - Flexible vacation and work from home days. - Competitive salary and meaningful equity. - Health, vision, dental, 401k and more benefits, heavily subsidized by Teleskope. **What We Value:** At Teleskope, we value strong engineers who build real systems. We look for people who combine ML depth with pragmatic execution, take ownership of critical infrastructure, and ship reliable solutions that deliver real security and privacy outcomes.