Prompt · Project Managers
Build Regulatory Update Chatbot
Use this when you want a conversational interface to deliver real-time regulatory updates and answer compliance questions.
How to use it
- Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
- Replace every {{placeholder}} with your own details, or let the AI ask you for them.
- Use the follow-ups below to go deeper.
Prompt
Role You are a chatbot designer and compliance expert who builds conversational assistants that keep project managers informed and compliant.
Context you provide
- {{industry}} — the sector for which the chatbot will track regulations.
- {{regulatory_sources}} — the official sources or feeds the chatbot should monitor.
- {{user_questions}} — the types of compliance questions the chatbot should answer.
- {{delivery_channels}} — where the chatbot will live (e.g., Slack, Teams, web).
Instructions
- Ask for any missing inputs before starting.
- Design a chatbot architecture that ingests regulatory updates from the specified sources and stores them in a searchable format.
- Define the chatbot's conversational flow: greeting, update summaries, and Q&A on specific regulations.
- Specify how the chatbot will deliver real-time notifications (e.g., push messages, daily digests) and on which channels.
- Outline the knowledge base structure and how the chatbot will retrieve accurate answers.
Output format A chatbot specification with sections: Purpose, Data Sources, Conversation Flow, Notification Logic, and Technical Requirements. Use bullet points and clear headings.
Guardrails
- Do not claim the chatbot can provide legal advice; frame it as informational.
- Flag any assumptions about the user's technical stack.
- Ensure the design prioritizes data accuracy and source citation.
Example Industry: healthcare; sources: CMS, FDA; user questions: 'What changed in HIPAA?'; delivery: Slack.
Follow-up prompts
- How can I test the chatbot's accuracy before deployment?
- What are the best practices for handling ambiguous user queries?
- Can you suggest a fallback for when the chatbot cannot find an answer?