The MongoDB AI Applications Program (MAAP) delivers swift enterprise-grade tools to build and deploy generative AI apps

MongoDB is already at the core of several well-known AI-powered solutions, as highlighted by its vast trove of case studies. Some case studies describe how MongoDB's solutions can enhance industry-standard approaches to critical use cases, including retail, healthcare, and cybersecurity. Yet others introduce a selection of existing companies that have successfully integrated MongoDB into their AI and ML stacks. The most recent of these case studies concerns contact center performance company Observe.AI, which has developed a 40 billion-parameter contact center large language model (LLM) jointly with one of the most accurate Generative AI engines to optimize customer service and sales interactions. The company leverages the MongoDB developer data platform to store the unstructured data associated with the millions of support touchpoints it processes to deliver useful insights to its customers.

The success stories of companies that have developed large-scale AI applications are a testament to MongoDB's value but also reveal an unaddressed gap in the enterprise ecosystem: not every organization has a dedicated department or enough private resources to unlock the value held by AI-powered automation and optimization. To address this issue and ensure that organizations of all sizes can quickly build and deploy secure, productivity-optimizing generative AI apps, MongoDB recently launched its AI Applications Program (MAAP). MAAP brings together "strategic advisory, professional services, and an integrated end-to-end technology stack from MongoDB and its partners." MongoDB's partners include industry-leading consultancies, and foundation model, cloud infrastructure, generative AI framework, and model hosting providers.

Cohere is among the first cohort of MAAP partners, and will make its Embed, Rerank and Command R models available via the Program. Cohere's model suite supports end-to-end retrieval-augmented generation (RAG) workflows: Embed and Rerank help optimize the retrieval stage of the RAG process by providing contextual understanding and relevance testing, respectively. Then, Command-R can perform a series of tasks on the retrieved data, including enterprise optimization essentials such as analysis and summarization. Customers leveraging Cohere's models as part of the MAAP can profit from their balanced and cost-efficient performance, as well as their multilingual capabilities. Additionally, Cohere supports on-premise and cloud deployments with all major providers.

Cohere joins a first cohort of announced partners including Anthropic, Anyscale, Fireworks AI, LangChain, LlamaIndex, Together AI, and more. The program is upcoming, and those interested can register for updates on the MAAP webpage.