Enterprise Knowledge Search
Swisscom needed to help banking users search and summarize a fast-growing library of more than 3,500 industry documents with better speed, relevance, and trust than general-purpose online AI tools could provide.
Reported impact
Swisscom stood up the first solution with about six days of development effort, moved from concept to prototype in days, improved the speed and relevance of banking-content searches and summaries, and created a reusable pattern for additional Swiss AI Platform services.
Swisscom built a retrieval-augmented generation service for e.foresight, its banking-industry think tank, by putting the entire document library on MongoDB Atlas and using MongoDB Vector Search to ground LLM responses. Reports, live-event recordings, graphics, and external studies are stored in Atlas as the single secure repository, then vectorized so the application can retrieve relevant segments and feed domain-specific context into large language models. The team went from first contact with MongoDB on February 6 to a working subset prototype by February 15, then production in early May after joint testing with ZHAW and internal operations handoff. Because the data stays in Switzerland and every answer can point back to cited source material, Swisscom can serve banks with faster, more accurate summaries while meeting the trust and sensitivity requirements of the financial sector.
Reports, event recordings, graphics, and external studies are stored in MongoDB Atlas, converted into embeddings, searched with MongoDB Vector Search, and the retrieved context is passed to LLMs so e.foresight users receive condensed, source-linked answers instead of manual document lists.
"Swisscom can then convert the unstructured data into vector embeddings and use MongoDB Vector Search on Atlas to find segments of relevant context to feed the LLMs."
"Total development effort for the first solution was around six days."
"With this project we’ve proved we can create something powerful using data held only in Switzerland."
"e.foresight has over 3,500 documents about trends in the banking industry in its library, with more being added each day."
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
MongoDB Sample Oct 8, 2026 Navisio AI Rapid adoption of Navisio AI's platform triggered severe latency spikes, elevated CPU and RAM consumption, and costly vertical scaling. As it prepares to scale to 1,000 by the end of the year, and eventually toward 5,000 users, the team is… MongoDB Atlas · MongoDB Atlas Vector Search MongoDB Sample Sep 25, 2026 Intellect Design Arena Intellect Design Arena needed a governed retrieval architecture that could turn fragmented ESG data into trustworthy enterprise AI answers for highly regulated financial institutions. An initial vector-only search achieved an accuracy rate of 71%. MongoDB Atlas · Purple Fabric MongoDB Sample Sep 9, 2026 Squary AI Squary AI needed to eliminate slow, outage-prone legal search infrastructure that combined MongoDB with OpenSearch and hurt document review speed, cost, and trust. At Quatrini Law Group, Squary AI analyzed more than 460,000 pages and saved over 400 attorney and paralegal hours. Amazon Bedrock · Claude · MongoDB Atlas 