Engineering teams globally are developing AI-centric applications or adding AI functionalities to existing products. The AI development environment is advancing, speeding up the prototyping of such applications. Despite this, deploying AI applications to production remains notably challenging. Contemporary AI stacks require LLMs, embeddings, vector searches, observability, novel caching layers, and continuous adaptation due to rapid changes. The data layer is increasingly both the foundation and the obstacle in AI app production.
MongoDB is extending beyond its main document database, evolving into a comprehensive AI-ready database platform with built-in features for operational data, search, real-time analytics, and AI-driven data retrieval. The company recently acquired Voyage AI to offer precise and cost-efficient embedding models and rerankers to its clients.
Fred Roma, a seasoned engineer and currently the SVP of Product and Engineering at MongoDB, speaks with Kevin Ball about the current state of AI application development, the importance of vector search and reranking, schema evolution in the LLM era, the Voyage AI acquisition, and how data platforms must advance to keep pace with AI’s rapid development.
Kevin Ball, or KBall, is the vice president of engineering at Mento and an independent engineering coach. He co-founded and was CTO for two companies, started the San Diego JavaScript meetup, and leads the AI inaction discussion group through Latent Space.
Please click here to see the transcript of this episode.