AI/ML Data Knowledge Graph Engineer

Where this role sits

This role builds the structured knowledge that collective intelligence reasons over. You own the KO (knowledge object) graph: the layer that turns a community's scattered expertise into connected, queryable knowledge the COGENT architecture can use.

You work where messy real-world data becomes trustworthy structure: ingesting, resolving, connecting, and modeling knowledge so reasoning has something solid to stand on.

You partner closely with neuro-symbolic AI and applied AI, and you are the reason the platform can answer questions that span a community's knowledge instead of isolated documents.
Why this role exists

Language models are fluent, but fluency is not knowledge. To reason over a community's expertise with rigor, the platform needs that expertise structured, connected, and trustworthy, not just retrieved as text.

Building a knowledge graph from real, fragmented sources is hard: entities to resolve, relationships to infer, quality to enforce, and provenance to preserve. The graph is only as good as the engineering behind it.

The AI/ML Data and KO Graph Engineer builds that foundation. You turn scattered knowledge into a graph the COGENT architecture can reason over, so members get answers grounded in their community's real expertise.
What you will own (Areas of Responsibility)

You hold seven areas of responsibility across the knowledge layer. Each one is yours to set direction on, build, and measure.
1. Knowledge graph engineering

• Build and maintain the KO graph that structures a community's knowledge for reasoning.
• Design schemas, ontologies, and relationships that reflect how expertise actually connects.
• Make the graph queryable, performant, and reliable at scale.

2. Ingestion and knowledge extraction

• Build pipelines that extract knowledge from documents, systems, and community sources into the graph.
• Turn unstructured and semi-structured content into structured knowledge objects.
• Keep the graph current as a community's knowledge changes.

3. Entity resolution and quality

• Resolve entities, deduplicate, and connect knowledge across fragmented sources.
• Enforce quality so members can trust what the graph tells them.
• Detect and handle conflicts and gaps in the knowledge.

4. Provenance and trust

• Preserve provenance so every piece of knowledge can be traced to its source.
• Build the structure that lets the platform show its work and earn member trust.
• Protect sensitive community knowledge with correct access and governance.

5. Serving knowledge to COGENT

• Partner with neuro-symbolic and applied AI to serve the graph into reasoning and retrieval.
• Shape the graph so it supports both symbolic reasoning and neural retrieval.
• Make knowledge access fast enough for production answers.

6. Data pipelines and platform

• Build the data pipelines and platform the knowledge layer depends on.
• Instrument the pipelines so quality and freshness can be measured.
• Turn recurring ingestion needs into reusable connectors.

7. Evaluation of knowledge quality

• Measure the quality, coverage, and freshness of the graph against what communities need.
• Build the evaluation that tells whether the knowledge layer is improving.
• Use evidence to steer where to invest next.

AI-augmented ways of working

You use AI to build the knowledge layer, from extraction and entity resolution to schema suggestions, while you own the correctness, structure, and trustworthiness of the graph.

The standard is human in partnership: AI accelerates the work, you own the judgment, the interpretation, and the call. The people who create the most value here are not the ones producing the most output. They are the ones turning evidence into clear, durable decisions.
What this role is not

To keep the boundary clear:

• This is not a reasoning-architecture role. You build the graph the COGENT architecture reasons over; neuro-symbolic AI owns the reasoning.
• This is not a general analytics role. You engineer a production knowledge graph, not dashboards and reports.
• This is not an ingestion-only role. You own structure, quality, provenance, and how knowledge serves reasoning.
• This is not a best-effort role. The graph is a production foundation members' trust depends on.

What success looks like

We measure this role on outcomes the team can see:

• Connected knowledge. A community's scattered expertise becomes a connected, queryable graph.
• Trustworthy answers. Quality and provenance let members trust and trace what the platform tells them.
• Fresh and current. The graph keeps pace with a community's changing knowledge.
• Reasoning-ready. The graph serves both symbolic reasoning and neural retrieval well.
• Reusable ingestion. New sources come online faster because ingestion is reusable.
• Measured quality. Coverage, quality, and freshness are measured and improving.

Who you are
Required qualifications

• Five or more years in data engineering, knowledge graph engineering, or a related field.
• Hands-on experience building and operating knowledge graphs or graph databases.
• Strong data pipeline engineering, including ingestion and transformation.
• Experience with entity resolution, deduplication, and data quality.
• Solid grounding in knowledge representation, ontologies, or schema design.
• Strong Python and SQL, plus graph query languages.
• Care for provenance, trust, and protection of sensitive data.

Preferred qualifications

• Experience serving graphs into retrieval or reasoning systems.
• Familiarity with neuro-symbolic AI and how structure supports reasoning.
• Experience with embeddings, vector search, and hybrid retrieval.
• Experience integrating CRM, AMS, or knowledge-base sources.
• Domain understanding of knowledge-intensive or professional communities.

How you work

• You name the real problem in the data before reaching for a structure.
• You care about quality, provenance, and trust as much as coverage.
• You build pipelines others can run and extend.
• You measure the knowledge layer honestly.
• You share reusable connectors and patterns.

Skills & Competencies

• Knowledge graph and ontology engineering.
• Ingestion, extraction, and transformation pipelines.
• Entity resolution, deduplication, and data quality.
• Provenance, governance, and protection of sensitive knowledge.
• Serving graphs into retrieval and reasoning.
• Evaluation of knowledge quality and coverage.
• Turning recurring ingestion into reusable capability.

Services & Tools Experience

• Graph databases (for example Neo4j-class systems) and graph query languages (Cypher, SPARQL, or GQL).
• Data pipeline and orchestration tools.
• Entity resolution and data-quality tooling.
• Vector databases and embedding models for hybrid retrieval.
• Python and SQL as primary languages.
• Cloud data platforms and storage.
• Building and serving the KO graph into the COGENT architecture and MINERVA.

Prior Experience & Background

• Prior data or knowledge graph engineering at a software or AI company.
• Experience building knowledge structures from messy, real-world sources.
• A track record of production data systems with quality and provenance.
• Experience supporting reasoning or retrieval systems is a plus.

Cross-functional partners

You work most closely with Neuro-Symbolic AI, Applied AI, Data Engineering, and Platform Engineering. You build the KO graph that the COGENT architecture reasons over inside MINERVA.
How we hire

We review every application, and we encourage you to apply even if you do not match every line above. Research shows that talented people, especially those from underrepresented communities, often hold back when they do not meet every qualification. If that is the only thing holding you back, apply anyway.
Compensation

Base Salary: $204,000 - $216,000 + early stage equity

Generous health and wellness benefits