AI Data Engineer - Data & Database Architect AI agents

The best AI agents start with quality data.

The opportunity

Genta helps small and mid-sized companies turn AI from scattered experiments into secure, measurable and operating AI agents. Our work combines practical implementation, human control, quality assurance and measurable ROI.

We are now looking for the person who can build the data foundation behind those agents: the databases, integrations, validation rules, audit trails and data-quality patterns that make AI outputs accurate, explainable and safe to use in daily operations.

This role works side by side with the AI Builder. Where the AI Builder turns workflows into agents, you make sure the agents can rely on the right data, from the right source, in the right structure, with the right level of confidence.

Role snapshot

Company: Genta

Role: AI Data Foundation Engineer / Data & Database Architect

Focus: Data models, databases, integrations, data quality and agent-ready data services

Core stack: PostgreSQL, SQL, Prisma, TypeScript and/or Python, APIs, Docker, Redis / BullMQ

Integration stack: ERP and business systems through REST, SOAP, OData, MCP, SQL Server and database-backed adapters

AI context: Support LangGraph / LangChain agents, n8n workflows, conversation memory, auditability and confidence scoring

Location: Copenhagen, Denmark

Engagement: Full-time

Reporting: Founder & Owner Bjørn Bahnsen

Why this role matters

An AI agent can classify an email, extract entities and draft an answer. But if the product ID is wrong, the order data is stale, the ESG value has no source, or the database cannot explain where a number came from, the agent cannot be trusted.

Your job is to make the data layer production-ready. That means designing clean source-of-truth structures, building robust integrations to customer systems, validating data before agents use it, and making every important answer traceable afterwards.

What you will build

• PostgreSQL data models for users, customer context, products, documents, conversations, agent checkpoints, whitelist rules, audit logs and operational metrics.

• Data pipelines and integration adapters that connect agents to ERP systems, product databases, ESG databases, spreadsheets, email metadata and customer-specific source systems.

• Agent-facing data services: clean APIs, structured query endpoints and tool schemas that AI agents can call safely and predictably.

• Data quality controls for completeness, freshness, duplication, entity matching, field validation, referential integrity and confidence scoring.

• Source-of-truth and lineage patterns so Genta and its customers can see which system produced which answer, when, and under what assumptions.

• Conversation memory and context storage that support follow-up questions without keeping more personal data than necessary.

• Caching, queueing and asynchronous processing patterns using Redis / BullMQ where response time, rate limits or PDF/report generation require it.

• Reusable data templates, schemas and implementation playbooks so Genta can build future agents faster and with higher quality.

Your responsibilities

• Work with customers, colleagues and development partners to understand the data behind each workflow before an agent is built.

• Map customer source systems, data owners, data quality risks and practical integration options.

• Design relational schemas, database migrations, indexes, constraints and access patterns that are simple, scalable and maintainable.

• Build ETL/ELT flows, validation jobs and integration adapters for APIs, SOAP, OData, SQL Server, files and legacy systems.

• Create data contracts and structured response schemas for agent tool calls, retrieval steps and confidence checks.

• Implement audit logs, retention rules, data minimisation, access controls and GDPR-aware handling of user and customer data.

• Help define quality metrics: data coverage, missing fields, stale records, unresolved entities, failed integrations and low-confidence answers.

• Document the data foundation in a way that both developers and business users can understand.

• Challenge bad data assumptions early. A polite no to unreliable data is better than a confident wrong answer from an agent.

What you bring

You are a strong match if you have worked with messy real-world business data and enjoy turning it into reliable, governed and usable systems.

Good to have

• Strong SQL and relational data modelling skills, especially PostgreSQL.

• Experience designing production databases with constraints, indexes, migrations, backup and operational discipline.

• Practical experience with data quality, validation, entity resolution and source-of-truth design.

• Ability to build integrations through APIs, files, databases and legacy systems.

• Backend capability in TypeScript / Node.js and/or Python.

• Security and privacy mindset: access control, auditability, GDPR, retention and data minimisation.

• Excellent communication: you can explain why data quality matters to non-technical decision makers.

Nice to have

• Prisma, Redis, BullMQ, Docker and EU-hosted cloud infrastructure.

• Microsoft Dynamics NAV / Business Central, SQL Server, SOAP or OData integrations.

• Experience supporting AI agents, LLM tool calling, RAG, vector search, embeddings or semantic retrieval.

• Familiarity with n8n, LangGraph, LangChain, MCP or LangSmith.

• Experience with product data, ESG data, CSRD, compliance, document generation or audit-heavy workflows.

• Data observability, lineage tools, metadata management or master data management.

• Danish language skills or experience with Danish SMEs.

This is not a dashboard-only role.

Business intelligence and reporting experience is valuable, but this role goes deeper. You will build the operational data foundation that AI agents use in real time: the structured data, validation layers, memory stores, integrations, logs and governance mechanisms that make agent decisions trustworthy.

You do not need to be a pure machine-learning researcher. You do need to understand that in agentic systems, data quality is product quality.

What makes you a great fit

• You care more about reliable systems than impressive demos.

• You can find structure in messy customer data without losing sight of the business problem.

• You think in schemas, contracts, ownership, lineage and failure modes.

• You are comfortable saying: the agent should not answer until the data is good enough.

• You enjoy working with both developers and business people, at eye level.

• You want to help make AI concrete, safe and useful in real companies.

Why join Genta

Genta is building a practical AI implementation company for businesses that want real value from AI without losing control, quality or human judgement. The market does not need more hype. It needs operating systems, trusted data and people who can translate technology into daily value.

You will have a central role in shaping the technical foundation of Genta's agent platform. Your work will determine whether agents can retrieve the right information, explain their answers, respect data boundaries and keep improving over time.

If AI agents are the visible interface, you will build the trusted data foundation underneath.

You are welcome to apply even if you have not used every tool in the stack. We care about deep data judgement, production discipline, curiosity and evidence that you can build systems AI agents can trust. But we care most about passion, curiosity and integrity.