Senior Engineer, Entity Graphs

About Firmable and the Role

Firmable stands as Asia Pacific's foremost B2B sales intelligence platform, supported by top investors and serving over 2,000 customers, with growing presence in the US. It compiles one of the most comprehensive datasets globally, detailing millions of companies, their personnel, and dynamic changes across hundreds of data sources combined into unified records accessible to humans and AI.

The Senior Engineer for Entity Graphs is responsible for architecting and owning the complex graph systems that underpin Firmable's comprehensive business knowledge. The role involves creating a hybrid data store that integrates graph databases, warehouses, and search technologies, optimized for AI agent traversal across a network of 15 million companies and 130 million people. This challenge demands innovative solutions to traditional graph retrieval problems, focusing on high confidence, efficiency, and scaling beyond conventional methods.

Main Responsibilities

• Design and maintain a large-scale hybrid architecture combining property graphs, data warehouses, and search indices to support comprehensive, consistent, and scalable data representation.
• Extend the graph schema continuously to accommodate emerging entity and relationship types without requiring migrations or compromising data integrity.
• Address structural challenges such as supernodes, traversal explosion, hot partitions, and unpredictable query costs unique to this massive graph.
• Integrate graph analytics like PageRank, community detection, and link prediction into the data as core corroborative inputs influencing confidence scores.
• Implement advanced entity resolution methods, including candidate generation, similarity scoring with various algorithms, clustering techniques, and threshold calibration to ensure high precision and recall across diverse and noisy data sources.
• Develop an agentic retrieval system that replaces traditional graph retrieval augmented generation (RAG) by enabling parametrized traversals and dynamic subgraph extraction tailored to query needs.
• Ensure safe and efficient translation of text queries to graph queries, managing traversal depth and query validation to control costs.
• Model entity lifecycle events such as mergers, splits, renames, and deregistration with bi-temporal data management and detect errors proactively before clients encounter them.

Key Skills and Experience Required

• At least 6 years designing and operating production data or graph systems, including ownership of architecture at scale.
• Expertise in AI-driven development workflows incorporating agentic coding, automated test generation, and continuous AI-driven code reviews.
• Hands-on experience with property graph databases like Neo4j or Memgraph, including expertise in Cypher query design, index planning, and managing traversal complexities.
• Proven capability to extend live graph schemas dynamically without service disruption or data inconsistency.
• Strong knowledge of graph algorithms (PageRank, community detection, shortest paths) and their performance at scales involving billions of edges.
• Deep understanding of hybrid storage architectures, balancing roles of graph stores, warehouses, and search engines with consistency guarantees.
• Expertise in machine-consumable retrieval systems, particularly graph RAG alternatives and query optimization techniques.
• Proficiency with search platforms like OpenSearch or Elasticsearch for candidate retrieval using fuzzy, phonetic, and kNN over embeddings.
• Experience managing cost and latency rigorously as key performance indicators for queries.
• Advanced skills in record linkage, master data management, and entity resolution at volumes beyond standard scalability thresholds.
• Familiarity with blocking, scoring, clustering, and calibration approaches using frameworks like Splink, Zingg, or Dedupe.
• Proven track record operating large-scale graph/linkage systems managing performance issues such as traversal blowup, hot partitions, memory ceilings, data skew, and costly full reprocessing.

Preferred Additional Expertise

• Experience with Snowflake data warehouse at scale, leveraging window functions, clustering, incremental merges, and cost control.
• Knowledge of distributed data processing frameworks such as Apache Spark or Snowpark including partitioning strategies and incremental processing.
• Familiarity with pipeline orchestration tools like dbt, Airflow, or Dagster.
• Experience managing registry data sources in production such as ASIC, ABR, NZBN, Companies House, or SEC EDGAR.
• Skills in resolving corporate hierarchies involving parent, subsidiary, trading names, and brand families.
• Understanding of inferred ownership attribution through analysis of shared IPs, multi-tenant infrastructure, and implied relationships.
• Exposure to vector search or learning-to-rank methods complementary to linkage strategies.

Working Environment and Culture

Firmable embraces a lean, fast-paced startup-to-scaleup setting where small cross-functional teams move with minimal bureaucracy and release weekly, moving towards daily cycles. Engineers have full ownership of their stacks, designing, building, shipping, and operating their solutions end-to-end. Flexibility in working hours complements the high expectations for delivering impactful results rapidly.

Firmable's AI-Native Development Philosophy

• All development processes are fully agentic, integrating AI for feature design, scaffold generation, and iterative coding under human guidance.
• Automated testing is driven by AI agents who craft context-aware tests, ensuring comprehensive coverage beyond manual effort.
• Continuous AI-powered code reviews maintain code quality and standards before human intervention.
• Real-time feedback loops with integrated tracing, evaluation, and observability systems from project inception.

Diversity and Equal Opportunity

Firmable values diversity and inclusion, recognizing that varied teams create superior products. The company maintains an equal opportunity hiring approach.