Agentic Data Architect

ProSearch; United States; On-site; USD 120,000 - 160,000; Full time

Job summary

ProSearch is seeking an Expert Agentic Data Architect to own the data flow architecture for agentic systems, ensuring access to accurate knowledge with proper governance across RAG, knowledge graphs, and agent memory layers.

You will design end-to-end architectures, set data quality standards, and drive the integration of vector and graph databases within a modern data stack. Strong governance experience and stakeholder collaboration are essential.

Qualifications

• 8+ years of data architecture, data engineering, or ML engineering experience.
• 2+ years with AI/ML or LLM based systems.
• Strong SQL and NoSQL knowledge across relational and non-relational patterns.
• Hands-on with designing and scaling RAG pipelines (chunking, embedding, indexing, hybrid search).
• Knowledge of vector and graph databases and their use in agentic retrieval.
• Experience with data governance frameworks and enterprise data security.
• Spec-driven development: data contracts, schemas, and pipeline specs.
• Proven ability to engage with business, legal, and data stakeholders to align on requirements.

Responsibilities

• Design end-to-end data architectures to support agentic workloads (ingestion, chunking, embeddings, vector stores).
• Define patterns for agent memory (short-term, episodic, semantic, procedural).
• Architect grounding strategies for agents over data lakes, warehouses, APIs, and documents.
• Establish data quality, freshness, and lineage standards for knowledge assets.
• Lead selection and integration of vector and graph databases and hybrid search systems.
• Collaborate with Data Governance and Legal to implement access controls and audit trails.
• Define evaluation frameworks for retrieval quality, hallucination rates, and knowledge currency.
• Collaborate with platform architects to integrate data services into the agent platform.
• Drive spec-driven development with data contracts and schema docs before implementation.
• Engage with business, data, and product stakeholders to translate knowledge requirements into architecture.

Skills

Data architecture

Data engineering

ML engineering

AI/ML systems

RAG pipelines

Vector databases

Graph databases

SQL

NoSQL

Data governance

Spec-driven development

Stakeholder collaboration

Tools

Vector databases

Graph databases

Data lakes / lakehouses

Cloud data warehouses

Streaming data platforms

Job description

At ProSearch, we believe that our clients are the best in the world, and they deserve a partner who is as committed to their growth and innovation as they are. We are a team of passionate problem-solvers, industry thought leaders, and trusted advisors who thrive on helping our clients achieve extraordinary outcomes. As we continue to expand our influence in the legal technology and eDiscovery sector, the Client Partnerships Advisor will play a critical role in supporting our strategic client relationships and advancing our mission of client success.

POSITION OVERVIEW

The Agentic Data Architect is the authority on how data flows into, through, and out of agentic systems. This role ensures that agents have access to the right knowledge, at the right fidelity, with appropriate governance — spanning retrieval-augmented generation (RAG), knowledge graphs, structured data access, and agent memory architecture.

RESPONSIBILITIES

• Design end-to-end data architectures that support agentic workloads: ingestion pipelines, chunking strategies, embedding models, vector stores, and retrieval layers.
• Define patterns for agent memory: short-term working memory, episodic memory, semantic memory, and procedural memory stores.
• Architect grounding strategies for agents operating over enterprise data sources (data lakes, warehouses, APIs, documents, real-time streams).
• Establish data quality, freshness, and lineage standards for knowledge assets consumed by agents.
• Lead the selection and integration of vector databases, graph databases, and hybrid search systems.
• Partner with Data Governance and Legal to implement access controls, PII redaction, and audit trails within agent data pipelines.
• Define evaluation frameworks for retrieval quality, hallucination rates, and knowledge currency.
• Collaborate with the Agentic Platform Architect to integrate data services into the agent platform.
• Drive spec-driven development for all data architecture work — producing data contracts, schema definitions, pipeline design docs, and retrieval system specs before implementation begins.
• Engage directly with business, data, and product stakeholders to understand knowledge requirements, data ownership boundaries, and governance constraints, translating them into concrete data architecture decisions.
• 8+ years of data architecture, data engineering, or ML engineering experience; 2+ years with AI/ML or LLM based systems.
• Deep expertise in modern data stack technologies: data lakes and lakehouses, cloud data warehouses, and streaming data platforms.
• Strong knowledge of traditional dimensional data architecture and modeling for analytical systems.
• Strong command of SQL and NoSQL data systems, including design trade-offs across relational and non relational patterns.
• Hands-on experience designing and scaling RAG pipelines (chunking, embedding, indexing, hybrid search, reranking).
• Strong understanding of vector and graph database technologies and their trade-offs for agentic retrieval use cases.
• Experience with data governance frameworks and enterprise data security.
• Demonstrated practice of spec-driven development: producing data contracts, schema docs, and pipeline design specs that precede implementation.
• Proven ability to engage directly with business, legal, and data stakeholders to align on knowledge requirements, data ownership, and governance constraints.

Preferred Qualifications

• Experience with knowledge graph construction and graph-based retrieval patterns.
• Background in NLP, information retrieval, or search engineering.
• Familiarity with structured data agents and natural language interfaces over relational and semantic data layers.

PHYSICAL REQUIREMENTS

• Repetitive movement of hands and fingers-typing and/or writing
• Have close visual acuity to perform an activity such as: viewing a computer terminal, extensive reading
• Occasional standing and walking

The duties and responsibilities described are not a comprehensive list. Additional tasks may be assigned to the employee from time to time, and the scope of the job may change as necessitated by business demands.

Compensation Range: $120,000 - $160,000 annually