back_office_ops · public · workflow

Vannevar Labs fine-tunes multilingual sentiment analysis model in 2 weeks with Databricks Mosaic AI, reducing latency by 75%

Vannevar Labs needed accurate multilingual sentiment analysis for defense intelligence work, but GPT-4 with prompt engineering achieved only 65% accuracy, was too expensive, and struggled with lower-resourced languages like Tagalog. GPU resource shortages and infrastructure management complexity also blocked the team from fine-tuning their own model.

How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · Ingest data from public sources
MCLI's data ingestion capabilities allow seamless, secure connection to Vannevar's datasets aggregated from multiple public sources.
Tools used
Databricks Mosaic AIDatabricks Model TrainingMCLIPython SDKMistralNVIDIA A10 Tensor Core GPUWeights & BiasesGPT-4
Outcome

The fine-tuned model achieved an F1 score of 76%, reduced latency by 75%, and was deployed within 2 weeks, enabling Vannevar Labs to process significantly more data more efficiently at lower cost.

What failed first

GPT-4 with prompt engineering failed to meet accuracy requirements for multilingual sentiment classification, achieving only 65% accuracy while being cost-prohibitive and inadequate for lower-resourced languages.

Results
Time saved76%
Volume65%
Cost replacedsignificant cost savings
Source

https://www.databricks.com/customers/vannevar-labs

How we source this →

Grounding & classification
Source type: vendor customer story
31 fields verified against source quotes.
document classificationsentiment analysissocial media postfailure mode describedmetric backednamed customerproduction runtime claimedsource backedtools describedworkflow describedpublic safety defenseaccuracy improvementcost reductioncycle time reductionthroughput increasevendor customer storyback office opsdocument to recordextract classify route