About
Nautilus needed to extend their team with AI and backend expertise to build an intelligent answer engine for dense ship technical documentation. UppLabs engineers embedded directly into Nautilus's product team, building a RAG-based platform ingesting maker manuals, class society records, and maintenance documentation into a governed, versioned knowledge base with natural language query capability.
Challenge, approach, and impact
Dense, Fragmented Maritime Documentation
Ship technical documentation including maker manuals, class society records, and maintenance logs existed across disconnected, unstructured sources, making it extremely difficult for engineers to quickly locate the specific technical answer they needed.
High-Stakes Accuracy Requirements
Incorrect or imprecise answers to technical maintenance questions carried real safety and compliance implications for vessel operations. The system needed retrieval accuracy that engineers could trust under operational pressure.
Governance Across Multiple Document Authorities
Documentation originated from multiple authoritative sources including manufacturers and class societies, each requiring proper versioning and governance to ensure engineers always referenced the current, approved version of any technical document.
Natural Language Access for Non-Technical Query Patterns
Engineers needed to query complex technical documentation using natural, conversational language rather than navigating rigid document hierarchies or relying on exact keyword matches.
How we built
Testimonials
Confidential C.
Nautilus
“UppLabs built a RAG pipeline that finally made our maritime technical documentation searchable in plain language. Their team understood the governance requirements around document versioning and authority that matter so much in our industry. Engineers can now get precise, trustworthy answers instead of digging through scattered manuals.“
Team structure
Client team
Confidential C
CEO
Project stakeholder
The client stakeholders at Nautilus were working closely with the team at UppLabs
Agency team
2 x Backend Developer
Production
1 x Machine Learning Engineer
Production