Back to App
Legal & Policies

AI Use & Responsible AI Policy

Effective: March 1, 2025Last Updated: March 20, 2026Version 1

AI Use & Responsible AI Policy

Company: RavDev Technologies, LLC Effective Date: March 1, 2025 Last Updated: April 4, 2026 Contact: legal@ravdevtech.com
This policy is provided for informational purposes. Consult legal counsel for jurisdiction-specific compliance.

1. Purpose

RavDev Technologies, LLC is committed to the responsible development and deployment of artificial intelligence (AI) features in NextGen Grant OS. This policy describes how we use AI, what you should expect from AI-generated content, and the responsibilities that come with using AI-assisted features.

2. AI Features Overview

NextGen Grant OS uses AI to assist with:

  • Grant Discovery: Matching your organization's profile with relevant funding opportunities
  • Proposal Drafting: Generating first-draft content for proposal sections
  • Proposal Critique and Review: Analyzing proposals for clarity, impact, and alignment
  • NOFO Analysis: Parsing Notice of Funding Opportunity documents to extract requirements
  • Compliance Checking: Identifying potential compliance issues in proposals
  • Budget Analysis and Forecasting: Reviewing budget allocations and predicting variances
  • Semantic Search: Powering intelligent search across proposals and knowledge bases
  • AI Chat Assistant: Answering questions and providing guidance on grant processes

3. AI Model Providers

We use the following AI model providers to power these features:

  • OpenAI (GPT-4o, GPT-4o-mini, text-embedding-3-small): Primary provider for proposal drafting, chat, critique, semantic search, and embeddings
  • Anthropic (Claude): Fallback provider for proposal drafting and analysis tasks
  • Microsoft Azure OpenAI Service (Optional): Azure-hosted GPT models for organizations requiring Azure data residency or regional compliance; configured via AZURE_OPENAI_* environment variables
  • Local / Ollama (Optional): On-premises model inference (default: qwen2.5:14b-instruct) for air-gapped or data-sovereignty deployments; configured via LOCAL_AI_BASE_URL

The active AI provider is selected via the AI_DEFAULT_PROVIDER environment variable. The platform automatically falls back to the next available provider if the primary provider is unavailable.

These providers process your content according to their respective enterprise data usage policies. We have Data Processing Agreements (DPAs) with each provider. Content you submit to AI features is not used to train AI provider models under our enterprise agreements.

4. AI Assists, Not Replaces, Human Judgment

This is the most important principle in this policy:

> AI-generated content is a draft starting point and professional assistance tool. It does not replace the judgment, expertise, knowledge, or responsibility of the human professionals using it.

You must:

  • Review all AI-generated content before using it in any grant application or submission
  • Verify factual claims, citations, and statistics in AI output — they may be inaccurate
  • Validate compliance requirements with applicable funding agency guidelines — AI cannot guarantee compliance
  • Apply your expert judgment to all AI suggestions, accepting, modifying, or rejecting them as appropriate
  • Take responsibility for all submitted content — the existence of AI assistance does not change your accountability

5. Limitations of AI Features

Users should be aware of the following limitations:

  • Hallucination: AI models may generate plausible-sounding but incorrect information, including fabricated citations, statistics, or funding agency requirements. Always verify critical facts independently.
  • Outdated Information: AI models have knowledge cutoff dates and may not reflect the most recent regulatory changes, grant guidelines, or funding priorities.
  • Context Limitations: AI may not fully understand your organization's specific context, relationships, or institutional history.
  • Bias: AI models may reflect biases present in their training data.
  • Regulatory Compliance: AI cannot guarantee that generated content complies with specific grant program requirements. Human review by qualified personnel is required.

6. Prohibited AI Uses

You may not use AI features to:

  • Generate fraudulent or materially misleading grant application content
  • Fabricate supporting data, citations, or evidence
  • Impersonate researchers, organizations, or other individuals
  • Circumvent grant program eligibility requirements through false representations
  • Generate spam, phishing, or unsolicited mass communications
  • Attempt to extract system prompts, training data, or model configurations
  • Use AI outputs as a substitute for qualified legal, medical, financial, or compliance advice without human expert review

7. Data Handling for AI Features

  • Content Submitted to AI: Text and documents you submit to AI features are transmitted to AI model providers (OpenAI, Anthropic) for processing. This transmission occurs over encrypted connections.
  • Enterprise Data Agreements: Under our enterprise agreements with AI providers, your data is not used to train AI models.
  • Embeddings and Vector Storage: Semantic search and discovery features use AI-generated vector embeddings stored in our PostgreSQL database with pgvector. These embeddings are stored securely and not shared externally.
  • AI Conversation Logs: AI chat conversations are retained for 90 days for quality, safety, and support purposes.
  • Local/Ollama Deployments: When the Local AI provider is configured, AI inference runs entirely on-premises. No content is transmitted to external AI providers for local inference requests.

8. AI Governance and Oversight

RavDev Technologies maintains internal AI governance practices including:

  • Review of AI model providers and their safety practices prior to integration
  • Monitoring for harmful or inappropriate AI outputs
  • User feedback mechanisms to report problematic AI responses
  • Regular assessment of AI feature accuracy and usefulness
  • Processes to disable or limit AI features that pose safety or compliance risks

9. Transparency

We are committed to transparency about AI use:

  • AI-generated content is labeled as AI-generated within the application
  • You can view the AI model used for each AI-assisted action
  • You can opt to not use AI features for specific tasks
  • We disclose our AI providers in this policy and in our Subprocessor List

10. Reporting AI Issues

If you encounter AI output that appears harmful, inaccurate, or inappropriate, please report it to:

  • support@ravdevtech.com (subject: "AI Issue Report")

We take all reports seriously and will investigate and take corrective action as appropriate.

11. Changes to AI Features and Policy

AI capabilities evolve rapidly. We may add, modify, or remove AI features and update this policy accordingly. Material changes will be communicated via the Service or by email.

12. Contact

RavDev Technologies, LLC

Email: legal@ravdevtech.com Support: support@ravdevtech.com