AI, connected to useful work.
An AI feature becomes useful when it fits a real task and has the right context. We integrate AI into software systems and workflows, from business-data chat in Purlin to editorial assistance and archive retrieval in KOHA’s CMS.
Talk about your projectWhat we
bring to it.
AI inside existing products
Connect model capabilities to the application, content and workflows your team works with.
RAG and archive access
Retrieval-augmented generation connects AI to relevant source material. KOHA’s CMS includes RAG and MCP access to its archive.
Chat and MCP integrations
Interfaces and tools that let people ask questions about the data in their systems, including Purlin’s business information.
Editorial and media tools
AI-assisted titles, descriptions, metadata and tags; text-to-speech, transcription, social media management and reel generation, as part of our KOHA work.
Considered from
the beginning.
- 01
Choose a concrete task
Define what the AI should help with, what information it needs and where a person reviews the result.
- 02
Connect the context
Plan access to relevant data, archive material and application tools.
- 03
Evaluate in the workflow
Check the integration against representative tasks and refine how results are presented and used.
A look at
our work.
A little
more detail.
What is RAG?
Retrieval-augmented generation gives an AI system relevant source material to use when answering a question. An editorial archive is one example: the system retrieves relevant material from the archive rather than relying only on a model’s general knowledge.
What is MCP used for?
The Model Context Protocol connects compatible AI clients to tools and data. We use MCP in products including Purlin, KOHA’s CMS and Grantpin.