Chatbot project overview

[priyadam]

1. Integrate Twilio Messaging API with Backend

Title: Integrate Twilio Messaging API for SMS/WhatsApp Support

Description:
Enable the backend to send and receive messages using Twilio’s Messaging API. This will allow users to interact with the chatbot via SMS or WhatsApp.

Tasks:

• Set up Twilio account and obtain API credentials.
• Implement webhook endpoint to receive incoming messages.
• Parse incoming messages and forward to chatbot backend.
• Send chatbot responses back via Twilio.

Acceptance Criteria:

• Users can send messages via SMS/WhatsApp and receive chatbot replies.
• Messages are logged and associated with a user session.

---

2. Connect LLM Chatbot Backend to Twilio Messaging Pipeline

Title: Connect LLM Chatbot to Twilio Message Flow

Description:
Bridge the Twilio webhook with the LLM-based chatbot backend (e.g., OpenAI or another LLM provider) to process user messages and return meaningful responses.

Tasks:

• Define message input/output formats for LLM interaction.
• Implement adapter layer to convert Twilio messages to LLM prompts.
• Maintain session context for each user.
• Handle error responses and timeouts gracefully.

Acceptance Criteria:

• Bot responds correctly to user queries over Twilio.
• Context is maintained per user session.

---

3. Setup Auth0 Authentication for Web and Twilio Users

Title: Implement Auth0 Authentication for LMS Users

Description:
Use Auth0 for secure authentication and session management. This ensures that only verified users can interact with the chatbot, both through the web interface and Twilio.

Tasks:

• Create Auth0 application and configure callback URLs.
• Implement login/logout on the LMS web interface using Auth0.
• Link Auth0 identity to phone number for Twilio users.
• Store user metadata (e.g., role, grant history) securely.

Acceptance Criteria:

• Web users can authenticate via Auth0.
• Twilio users are matched to their Auth0 identity using phone number verification or OTP.
• JWT tokens are used to authorize access to chatbot API.

---

4. Create User Onboarding Flow with Auth0 and Twilio

Title: Build Onboarding Flow with Auth0 and Twilio OTP

Description:
Enable new users to register and authenticate using phone numbers, backed by Auth0. Use Twilio for OTP delivery.

Tasks:

• Trigger Auth0 phone-based login flow.
• Use Twilio to send OTP codes.
• Verify and store user profile after login.
• Link user to appropriate LMS role (e.g., applicant, mentor).

Acceptance Criteria:

• New users can register via Twilio using phone number.
• User sessions are tracked and secured via Auth0.

---

5. Implement Access Control in Chatbot Based on Auth0 Roles

Title: Add Role-Based Access Control (RBAC) in Chatbot Using Auth0

Description:
Ensure chatbot functionality respects user roles (e.g., applicant vs. reviewer). Roles are managed in Auth0.

Tasks:

• Define roles and permissions in Auth0 dashboard.
• Decode JWT tokens and enforce role-based logic in chatbot.
• Limit access to certain prompts or features based on role.

Acceptance Criteria:

• Users only access content relevant to their role.
• Unauthorized actions are blocked with appropriate messages.

---

6. Sync User Accounts Between LMS and Chatbot System

Title: Implement User Sync Between Main LMS and Chatbot System

Description:
Ensure users in the main LMS are mirrored in the chatbot system. This includes importing user roles, profiles, and access credentials.

Tasks:

• Create data sync service using LMS API or database access.
• Map LMS user roles to Auth0 roles.
• Implement periodic sync job or webhook triggers on LMS events.
• Handle user deactivation or permission revocation.

Acceptance Criteria:

• LMS users appear in the chatbot system automatically.
• Role changes in LMS reflect in chatbot permissions.

---

7. Enable SSO with Existing LMS via Auth0

Title: Enable Single Sign-On (SSO) Between Chatbot and LMS via Auth0

Description:
Allow users to authenticate into both systems using the same credentials via SSO. If the LMS supports SAML, OIDC, or LDAP, integrate it with Auth0.

Tasks:

• Configure LMS as a SAML or OIDC identity provider in Auth0.
• Test login flow from LMS to chatbot and vice versa.
• Maintain a consistent session across systems.

Acceptance Criteria:

• Users log in once and can access both the LMS and chatbot.
• Auth0 federates identity across both platforms.

---

8. Fetch Grant Learning Modules from LMS into Chatbot

Title: Fetch Grant Learning Modules from LMS for Chatbot Access

Description:
Pull structured grant-related content from the LMS into the chatbot for contextual delivery during conversation.

Tasks:

• Define API interface for retrieving module data (e.g., lessons, quizzes).
• Add chatbot intents to deliver this data based on user queries.
• Cache or sync content periodically to reduce LMS API calls.

Acceptance Criteria:

• Users can ask the chatbot for specific LMS content (e.g., “Explain part B of the NIH application”).
• LMS content is displayed contextually in chatbot responses.

---

9. Submit Progress or Quiz Results Back to LMS

Title: Sync Chatbot Quiz Results or Interactions Back to LMS

Description:
Allow the chatbot to update user progress, scores, or completion status in the main LMS after relevant interactions.

Tasks:

• Define webhook or API endpoints in LMS for receiving chatbot updates.
• Track completed modules or practice exercises within chatbot.
• Push this data to LMS in proper format (e.g., SCORM/xAPI if needed).

Acceptance Criteria:

• Chatbot interactions are reflected in LMS progress tracking.
• LMS reports include chatbot-originated activities.

---

10. Unified Dashboard for Admins to Monitor User Activity

Title: Create Admin Dashboard Integrating Chatbot and LMS Analytics

Description:
Provide LMS admins with a dashboard that shows user activity across both the LMS and the chatbot system.

Tasks:

• Build a backend service to aggregate data from both systems.
• Include metrics like login times, module access, and chatbot usage.
• Provide exportable reports or visual dashboards.

Acceptance Criteria:

• Admins can view combined engagement metrics.
• Reports can filter by user, module, or interaction type.

---

Above are the notes from ChatGPT
Further steps:

• Check the progress for LMS application with David
• Need to connect Auth0 with backend
• Auth0 and Twilio credentials already done in last cohort and some basic implementation which we can bring to this LMS application (Shared by Hui)
• Create issues based on notes shared by ChatGPT above
• Research AI agents that we could use