PermitAI is building a suite of AI-native web applications to support federal agencies in efficiently managing, analyzing, drafting, disseminating, and visualizing environmental review and permitting data. PermitAI Applications will enable agencies to make faster, more informed permitting decisions, reducing delays and costs. For now, these tools are only available for beta testing by federal government users.  

SearchNEPA

SearchNEPA, helps permitting professionals find the documents they need and get verifiable answers from them, without combing through hundreds of thousands of PDFs by hand.

Every document in its database is tagged with metadata such as project, process, and document type, so users can filter precisely. Because projects are geolocated, users can also explore them on an interactive map, finding projects by place and seeing how they cluster geographically. Advanced search capabilities surface the most relevant documents, and AI assistants handle both quick, pointed questions and broader, complex questions that delivers trends and analytics derived from historical permitting documents. Answers are designed to be easily verifiable, so you can check each claim against its source.

SearchNEPA is PermitAI's first application. Launched in late 2024, it has more than 600 onboarded users and it is available for beta testing by federal government users from the Department of Energy, the Department of Agriculture, the Department of the Interior, National Institute of Standards and Technology, and the U.S. Army Corps of Engineers.
 

CommentNEPA

CommentNEPA is an AI-enabled tool designed to streamline one of the most time-intensive steps in the public comment review process: organizing, analyzing, and summarizing submitted comments.

The application processes correspondence to identify duplicate submissions and distinguish unique text within form letters. It then extracts individual comments, organizes them into user-defined or AI-generated categories, and generates summaries to support review and analysis. Results can be exported to Excel for additional processing, documentation, or integration into existing workflows.

A key feature of CommentNEPA is its auditable, interactive refinement framework, which allows users to review, refine, and adjust AI-generated outputs based on subject matter expert input. Users can re-bin extracted comments, modify comment groupings, and regenerate summaries to better align results with project-specific needs and expert judgment.

CommentNEPA is available for use by federal government users.

WriteNEPA

WriteNEPA is a National Environmental Policy Act drafting assistant built to assist subject matter experts with initial drafts of content for EIS and EAs. The application uses a unique workflow that combines defined style, tone, and formatting conventions with action-specific data for a new project to create a robust draft that comprises information from environmental reports, websites, federal datasets, and more. The end result is a fully auditable, cohesive first draft that a subject matter expert can quickly verify and shape into a publishable document. 

EngageNEPA

EngageNEPA tracks NEPA actions and analysis of comments. Users can search for notices and meetings by keyword, agency, time periods, and geographic location. Once located, users can read summary of the notice, navigate to the ground-truth Federal Register pages, or go to regulations.gov to leave a comment.
 

PermitCE

PermitCE is a web-based chat tool that guides users through preparing a categorical exclusion (CE) evaluation for a project under review. The tool helps users draft project descriptions, review recommendations for applicable CEs, compare recommendations with similar CE determinations, and check project descriptions and locations against agency-specific requirements (e.g., extraordinary circumstances) using available environmental data.

PermitReg

PermitReg reviews statutory and regulatory citations in project documents for alignment with authoritative federal government sources — including the United States Code (USC), Code of Federal Regulations (CFR), Federal Register notices, and Executive Orders. It provides rapid feedback indicating whether each citation appears correct, or if it may be incomplete, unsupported, or potentially outdated.

PermitLex

PermitLEX helps NEPA practitioners find relevant court decisions and understand what courts have focused on, without reading hundreds of opinions one by one. 

Search by case name, docket number, agency, or keyword, or filter its database of 857 federal NEPA decisions across 102 agencies. Each case is tagged with metadata such as court, ruling date, state, NEPA document type, challenged issues, and remedy, so users can narrow to cases that resemble their own project. For each decision, AI-assisted analysis extracts the issues raised, the court's reasoning, and the deficiencies identified. Where the underlying NEPA documents are in SearchNEPA, cases link directly to them.

PermitLEX is a decision-support tool, not a legal decision-maker. It supports professional review and does not replace it.

NEPA-MCP

NEPA-MCP lets AI agents pull from the federal environmental data used in early NEPA screening, so practitioners can ask about a project site in plain language instead of working through agency web tools one at a time. 

Built on the Model Context Protocol (MCP), it provides more than 20 servers covering species and critical habitat, wetlands, floodplains, protected lands, historic properties, air quality monitoring, socioeconomic indicators, tribal geographic areas for consultation screening, and the eCFR and Federal Register. Given a project location or area, an assistant can query these sources and compose the results into an interactive map or a GeoJSON export for QGIS or ArcGIS. Each map reports which layers loaded, came back empty, or failed, so gaps in source coverage stay visible. It works with Claude Code, VS Code, and Codex. 

NEPA-MCP is open source and is a screening and research aid, not a legal or agency determination, and material findings should be confirmed against authoritative records.