AI Engineer

Datifact is hiring an AI Engineer to build the data architecture, ontologies, and LLM systems behind client deployments. The role spans creating data foundations and semantic models, developing production LLM applications and agents, and shaping the platform architecture and engineering practices.

Responsibilities:

• Design and build client data foundations: ingestion pipelines, warehouse and lakehouse models, and integrations with ERP, CRM, accounting and operational systems (QuickBooks Online, Salesforce, Snowflake and similar)
• Own data modeling standards, schema design and data quality across multi-tenant environments
• Build ingestion for unstructured sources (documents, contracts, email, call notes) and turn them into structured, queryable context
• Make security, tenant isolation and access control first-class parts of every architecture
• Model how a business actually works: entities, relationships, metrics, rules and processes, expressed as an ontology an AI system can reason over
• Build and maintain reusable domain packs (Finance and Operations, Procurement, Sales, Customer Success) and industry overlays that accelerate every new deployment
• Define a governed semantic layer so every metric means one thing, everywhere, for every user and agent
• Work directly with client finance and operations leaders to translate how they run the business into models
• Build production LLM applications and agents that retrieve, reason and act over client data and context
• Design retrieval systems (vector, graph and hybrid), prompt and context strategies, and tool-calling workflows
• Build evaluation harnesses that measure accuracy, grounding and reliability, and use them to drive every release
• Optimize for cost, latency and quality across frontier and small language models
• Ship to production and own what ships: monitoring, observability and iteration based on real usage
• Partner with the CTO on platform architecture and technical roadmap
• Work alongside solutions engineering and client delivery on discovery, builds and managed services
• Document patterns so the team can repeat them, and help shape engineering culture as Datifact grows