Data Scientist

Terzo

9d•Remote

About The Position

As a Data Scientist on our Applied Research team, you will build the intelligent systems that create the data our customers depend on. You will design extraction and classification models that process enterprise-scale document corpora, build and evolve the entity resolution and signal detection layers powering the Commercial Graph and Financial Graph, and define how AI capabilities surface as recommendations, agents, and search across the platform. You will own the models, pipelines, and graph structures that are the product — working directly with engineering, product, and customers on problems where a single clause can represent tens of millions of dollars of exposure and where model accuracy has a contractual SLA.

Requirements

5+ years of experience in data science, applied ML, or AI research with production-shipped systems, not just notebooks and prototypes
Strong statistical foundations and the ability to define and evaluate success metrics for AI systems including precision, recall, coverage, latency, not just accuracy
Deep experience building NLP, NLU, or document understanding models that operate on messy, real-world unstructured data at scale
Strong intuition for entity resolution, knowledge graph construction, or graph-based modeling and you've thought seriously about how to connect fragmented data into structured, queryable representations
Hands-on proficiency in Python and modern AI frameworks, with experience deploying models into production pipelines
Comfort with information extraction, classification, and retrieval-augmented generation patterns applied to real enterprise workloads
A track record of working cross-functionally with engineering and product to shape what gets built, not just executing on handed-down specs
Clear, structured communication where you can explain a model decision to a PM, defend an architectural choice to a staff engineer, and present results to leadership without hiding behind jargon
High ownership mentality where you treat model quality, pipeline reliability, and customer outcomes as your responsibility

Nice To Haves

Experience building or evolving knowledge graphs, commercial ontologies, or financial data models in enterprise contexts
Prior work on document AI, OCR pipelines, or hybrid extraction systems combining rule-based and learned approaches
Exposure to AI agent architectures, tool-use patterns, or autonomous reasoning systems in production
Background in procurement, contract management, spend analytics, or financial operations domains
Experience with evaluation frameworks for AI systems (RAGAS, custom eval harnesses, human-in-the-loop QA pipelines)
Familiarity with distributed data platforms, event-driven architectures, or streaming systems (Ray, Kafka, Azure Service Bus)
Prior work at a high-growth startup or enterprise AI company
An MS or PhD in a quantitative field

Responsibilities

Design extraction and classification models that process enterprise-scale document corpora.
Build and evolve the entity resolution and signal detection layers powering the Commercial Graph and Financial Graph.
Define how AI capabilities surface as recommendations, agents, and search across the platform.
Own the models, pipelines, and graph structures that are the product.
Work directly with engineering, product, and customers on problems where a single clause can represent tens of millions of dollars of exposure and where model accuracy has a contractual SLA.