Senior Associate Data Modeler

Are you looking for an exciting opportunity to join a dynamic and growing team in a fast paced and challenging area? This is a unique opportunity for you to work in our team to partner with the Business to provide a comprehensive view.

As part of Risk Management and Compliance, you are at the center of keeping the firm strong and resilient. You help JPMorganChase grow its business in a responsible way by anticipating new and emerging risks and using your expert judgement to solve real-world challenges that impact our company, customers and communities. Our culture in Risk Management and Compliance is all about thinking outside the box, challenging the status quo and striving to be best-in-class.

As the Data Analyst & Modeler in the Office of Regulation and Document Governance (ORDG), you will be the hands-on practitioner who takes ORDG’s policy, standard, procedure and regulatory content apart at the atomic level — clauses, obligations, controls, defined terms, cross-references and their relationships — and turns it into structured, analyzable, machine-readable data. Reporting to ORDG’s Data Governance & Data Product Lead and working alongside the Data Modeler & Business Analyst and the Data Analytics Lead, you will use data-science techniques (statistical analysis, NLP, entity and relationship extraction, clustering, similarity and graph methods) to profile the estate, prototype logical and physical models, validate them against real content, and generate the evidence base that governance and product decisions are made on. You will be equally comfortable in a notebook and in a modelling tool, and will pair with Data Science, Technology and consuming platforms (FPSP, DOGMA, LRCM and the RM&C Bulletin Tool) to make sure what we design at the whiteboard actually holds up on the real corpus.

Job responsibilities

Decompose ORDG’s document corpus at the atomic level — clauses, obligations, controls, defined terms, cross-references and citations — and build the entity and relationship views that describe how policy content actually connects.

Profile the governed data estate — completeness, cardinality, duplication, drift, orphaned references and structural anomalies — and surface findings that inform data-product design, remediation and prioritisation.

Apply NLP and text-analytics techniques (entity extraction, relationship extraction, similarity, clustering, topic modelling) to reveal structure inside unstructured policy content and translate it into candidate data structures.

Build and maintain a knowledge-graph view of the corpus — obligations, controls, defined terms, policies, procedures and their citations and dependencies — to make cross-document relationships explicit and queryable, and to underpin retrieval, impact analysis and AI grounding.

Prototype and iterate conceptual, logical and physical data models for ORDG’s governed data products in partnership with the Data Modeller & Business Analyst — including entity/attribute definitions, cardinality, hierarchy and lineage.

Validate proposed models against real content at scale — loading representative samples, testing rules, measuring coverage and edge-case behaviour — and feed evidence back into design decisions before physical build.

Build reference data structures for canonical taxonomies (obligations, controls, defined terms, jurisdictions) and maintain the mappings between source content and canonical form.

Support the Data Modeller & Business Analyst and Technology partners on physical implementation — data contracts, schema binding, transformation logic, validation rules — and hold artefacts accurate as the model evolves.

Use Python and SQL as everyday tools — pandas, scikit-learn, spaCy / NLP libraries, graph libraries, notebook-based analysis — to interrogate content, prototype models and generate evidence.

Design and run experiments to answer specific product and governance questions (e.g. ‘how many obligations overlap across two policies?’, ‘what is the citation graph for this domain?’, ‘where does a controlled term drift?’) and communicate results in a way non-specialists can act on.

Partner with Data Science on retrieval quality, extraction accuracy and model evaluation — including labelling, gold-set curation and error analysis on AI-generated structured output.

Produce the analytical artefacts that inform governance forum decisions — profile reports, coverage analyses, similarity maps, model change impact assessments — with a clear audit trail from raw content to conclusion.

Contribute to the ORDG analytics estate in partnership with the Data Analytics Lead — including estate-health signals, KPI/KRI feeds and content-level telemetry — as governed inputs into the unified analytics platform.

Support user-acceptance testing and data-quality validation for new or amended data products; capture defects, quantify impact, and re-baseline artefacts once accepted.

Work across time zones and delivery streams from Bangalore — pairing daily with governance, modelling, analytics, Technology and Data Science partners in the U.K., U.S. and India — and holding artefacts as the single source of truth across the handover.

Contribute to backlog refinement across the Define & Govern Data Products workstream: break down asks into modelling and analytical work items, estimate effort, and sequence delivery with the Data & AI Execution Lead.

Required qualifications, capabilities, and skills

Experience in a hands-on data analyst, data scientist or data modeler role — including entity/relationship modelling, data profiling and analysis on real business content.

Working knowledge of Python and SQL for data analysis — including pandas, notebook-based workflows, and one or more of scikit-learn, spaCy / an NLP library, or graph libraries (e.g. NetworkX, Neo4j).

Experience applying data-science techniques (statistical analysis, similarity, clustering, entity/relationship extraction, topic modelling) to unstructured or semi-structured text content.

Familiarity with conceptual, logical and physical data modelling — including normalization, canonical modelling, cardinality and lineage — and comfort iterating models against real content.

Comfortable working across Technology, Data Science, Product Owners and business SMEs, including U.S., U.K. and India partners; able to hold rigour on definitions while remaining pragmatic on sequencing.

Familiarity with AI tools, inclusive of LLM Suite, Chat GPT and/or Claude, and AI-assisted development tools such as GitHub Copilot, and an understanding of how governed structured data underpins safe AI consumption.

Preferred qualifications, capabilities, and skills

Prior experience analyzing or modelling policy, regulatory or control-related content — including obligations, controls, standards or procedure text.

Working knowledge of knowledge graphs and graph analytics — property graphs and/or RDF, Cypher / SPARQL, ontology and taxonomy design — for representing entity relationships, citation networks and controlled vocabularies.

Familiarity with modern data-catalog and lineage tooling (e.g., Collibra, Alation, Immuta) and modelling tools (e.g., ER/Studio, Power Designer, or equivalent).

Exposure to LLM-based extraction and evaluation techniques — including gold-set curation, error analysis, and prompt or fine-tuning approaches for structured output.

Familiarity with JPMorganChase data standards, CDAO frameworks and firmwide governance operating models.

Experience with common delivery and reporting tooling (e.g., Jira / Confluence and dashboard tools such as Power BI, Tableau or ThoughtSpot); strong Excel skills.

Experience operating in a controlled environment with governance processes, audit interactions and documentation rigor.