LIGER™ Generative AI Platform

LIGER™ is LMI’s secure, scalable, commercial-off-the-shelf (COTS) generative AI (gen AI) platform for government enterprise, with collaborative workflows designed to maximize gen AI’s productivity impact on an agency’s core mission. Cloud-agnostic, model-agnostic, and fully containerized, it is designed to address the unique security needs of the federal government to handle sensitive information. LMI built LIGER™ to help agency employees spend less time on the bureaucracy, and more time focused on delivering their core mission.

LIGER™ specializes in processing text and images from various file formats, including PDF, Microsoft Word, and PowerPoint files. Its Retrieval Augmented Generation (RAG) architecture is model-agnostic, meaning it can be used with any large language model (LLM). LMI has the flexibility to choose a model that best matches the available compute environment and desired capabilities for each customer deployment. This adaptability has allowed LMI to successfully deploy LIGER™ on air-gapped networks without access to cloud resources and even on a standalone laptop for field use without GPU or internet access. LIGER's design enables it to accurately retrieve information with minimal hallucinations, provide simple-to-use citations, and minimize compute cost/ token consumption with actionable and transparent cost management tools.

Benefits of using LIGER™ include an intuitive user interface designed around enabling humans to streamline work processes, the ability to securely deploy one or more open-source or proprietary LLMs, and elimination of the need to train LLMs on any government data – preventing unintentional data leaks. LMI achieves this through custom prompts, examples, and user-configurable workflows that instruct the model on what to generate, the desired format, and where to find the required information to complete tasks.

Pricing (effective Jan 1, 2025 through Jun 30, 2025) 

The base price for a pilot includes a CLIN for the deployment and support of LIGER™ and an ODC CLIN for the compute resources required to operate LIGER™. LMI does not charge a license fee per user for LIGER™, the customer can onboard as many users as desired at the same fixed cost, however the compute resources available for a pilot are capped at a GPU instance. If usage expands beyond the scale of the pilot deployment, then the customer will need to expand to a full deployment to maintain the quality of the user experience. All prices listed assume deployment in an unclassified environment.

Table 1: Pilot Deployment (Unclassified Environment)

Additionally, optional teams can be engaged to support custom development and scaling efforts. This allows LMI to help build more specific functionality required for a single customer’s use case(s). 

Table 2: Custom Development (Unclassified Environment) - Optional

At any time, the customer can convert the pilot into a full enterprise capability. This will allow the tool to expand from a single instance and scale the resources to meet the demands of as many users as required.

Table 3: Enterprise Deployment (Unclassified Environment)