Multi-Source Data Architecture

• Work with systems handling multiple write paths: external providers, LLM hygiene agents, and customer-claimed edits
• Define standards for data versioning, lineage, and observability across pipelines

Entity Lifecycle & Master Data Management

• Handle entity lifecycle complexity: mergers, acquisitions, spin-offs, rebranding, and temporal relationship changes
• Design entity resolution systems using deterministic blocking (fuzzy matching, location) combined with LLM-based evaluation for match decisions
• Build confidence scoring models and surface low-confidence cases for human review

Machine Learning & Matching Systems

• Work with embeddings infrastructure: vector generation, retrieval optimization, and quality measurement
• Optimize semantic search pipelines including embedding strategies, namespace design, and reranking
• Establish evaluation frameworks to measure model performance against human judgment

Collaboration & Team Development

• Educate and mentor the engineering team on data best practices, patterns, and common pitfalls
• Lead continuous improvement of the data infrastructure roadmap

Relationship & Graph Modeling

• Design data models for complex relationships: parent/subsidiary hierarchies, PE firm → portfolio company chains
• Evaluate and implement graph query capabilities (Apache AGE, Neo4j, or optimized Postgres patterns) for relationship traversal that semantic search cannot address

Data Quality, Testing & Operations

• Build quality-control layers including confidence scoring, human-in-the-loop validation, and automated anomaly detection
• Implement testing strategies including data contracts, pipeline unit tests, and integration testing
• Build proactive monitoring, alerting, and runbooks for data health issues
• Ensure compliance with data governance, privacy, and security standards

• Prior experience with Machine Learning algorithms / semantic search
• Prior experience with entity resolution or master data management — you understand why matching company records is fundamentally hard
• Familiarity with graph databases or graph query patterns (Neo4j, Apache AGE, recursive CTEs) for complex entity relationships
• Experience with event sourcing or append-only architectures for audit trails and data replay
• Background in investment data, market intelligence, or deal sourcing platforms
• Familiarity with agent orchestration tools (LangChain, LlamaIndex) and data quality frameworks (dbt, Great Expectations)
• Experience as an early/first data hire at a startup
• Experience designing or integrating AI/LLM agents for data enrichment with structured AI → JSON → database pipelines including error recovery and monitoring
• Understanding of prompt engineering, MCP Servers, function calling, and embedding-based retrieval