Donovan - 2026 Pricing, Features, Reviews & Alternatives

Donovan is an AI-powered decision-making platform built for U.S. federal government agencies, the Department of Defense, the Intelligence Community, and military operators at any classification level. The platform ingests structured and unstructured data and exposes it through natural language interaction, allowing operators and analysts to understand, plan, and act without manual data processing. Deployed across unclassified, classified, and air-gapped networks, Donovan holds a DISA Impact Level 4 (IL4) provisional authorization and a FedRAMP High in-process designation from the Joint Authorization Board. It is also available on AWS Marketplace and the AWS Intelligence Community Marketplace.

Built on the Scale GenAI Platform (SGP), Donovan functions as an agent ecosystem where mission-tailored AI agents are configured, deployed, and managed through a no-code Agent Factory interface. Core capabilities include Retrieval-Augmented Generation (RAG) for document querying with cited, source-linked responses; Text-to-API for translating natural language into live API calls against connected systems; Geospatial Chat combining geographic map filtering with large language model (LLM) inference; and Report Generation using customizable templates for mission-relevant documents and courses of action (COAs). The platform supports leading commercial foundation models alongside custom fine-tuned models, including a Defense LLM built on Meta Llama 3 and fine-tuned with DoD domain-specific knowledge and ODNI writing style guidelines, available exclusively in controlled U.S. government environments.

Donovan's model-agnostic and infrastructure-agnostic architecture supports flexible model selection, fine-tuning via Scale Data Engine, and integration with existing applications and tools, including Anduril Lattice for command and control. Large multimodal model (LMM) support extends use cases to geospatial imagery object identification and map interpretation. A built-in Model Evaluation framework lets teams compare model responses and performance metrics using testing frameworks aligned to DoD AI readiness goals, supporting informed model governance across the organization.