Case Study: Tenali AI achieves sub-second in-call intelligence with MongoDB Atlas and Voyage AI

Tenali AI cuts retrieval latency by 67% with MongoDB

Tenali AI, a computer software company, faced the challenge of eliminating the "I'll get back to you" moment on live enterprise sales calls. Their platform needed to instantly retrieve accurate answers from a vast and disorganized knowledge base during a conversation, requiring sub-second latency that traditional databases and fragmented stacks could not provide.

To solve this, Tenali AI implemented a solution using MongoDB Atlas with its Vector Search capability and integrated Voyage AI's embedding models. This unified data platform provided the necessary speed and accuracy, reducing retrieval latency by 67% from 300ms to 100ms. The result was that sales reps could answer technical questions instantly on live calls, which increased their in-call usage of the platform's answers by over 40% and helped maintain crucial deal momentum.