Lamini announces it raised $25 million to accelerate its Enterprise AI platform

Lamini is an enterprise AI platform that enables software teams within organizations to develop LLMs customized with their proprietary data locally. Lamini calls its approach 'Expert AI' to set it apart from General AI and highlights that Expert AI has deeper domain-specific capabilities and has to be built using a specialized process. The company has raised $25 million from Series A leaders Amplify Partners, First Round Capital, who led the Seed funding round, and a lineup of additional investors including Andrew Ng, Andrej Karpathy, Bernard Arnault, Pierre Lamond, Sarah Guo, Dylan Field, Lip-Bu Tan, Drew Houston, AMD Ventures, among others.

According to the company, following the approach that most individual users would take to profit from generative AI is detrimental for enterprises because of foundational models' propensity to hallucinate, the fact that many open-source libraries are research-focused, compliance and security requirements, and the lack of certainty on the ROI. Lamini has optimized its LLM stack to address every one of these concerns. As a result, enterprises can build personalized LLMs using their proprietary data to turn foundation models into enterprise-grade Expert AI deployments. Moreover, the Lamini platform also enables this process to develop in several environments: on the cloud, on-premise in air-gapped environments, or across the offerings of several compute vendors.

Lamini's solutions include memory tuning for reducing hallucinations, faster time to market and scalability for engineering teams, secure and flexible deployments across cloud and on-premise environments, and higher initial ROI through capital-efficient Expert AI development. The company has partnered with AMD, Snowflake, Databricks, and others, and is hiring for various roles to accelerate development and support enterprises at scale. In addition, the fast-growing company has unlocked milestones, such as developing its LLM Photographic Memory Evaluation Suite, Multi-node LLM Training, Guaranteed JSON Output, and Parameter-Efficient Fine-Tuning (PEFT). During this time, Lamini has also partnered with some of the biggest industry leaders in AI, including AMD, Snowflake, Databricks, Nutanix, Meta, Mistral, and others.