SharePoint RAG Chatbot Development
Budget: $8 – $15 USD
I need a Retrieval-Augmented Generation (RAG) chatbot embedded directly in our Microsoft SharePoint Online intranet so colleagues can surface the right content, understand it quickly, and act on data-driven insights without leaving the site.
Core interactions
• Content discovery – when someone asks a question, the bot should crawl or query published intranet pages, list documents, and any connected knowledge bases, then return the most relevant pages, files, or list items.
• User support – conversational answers must be generated in natural language, summarising or explaining information from the source page so users know immediately why that item was recommended.
• Data analytics – capture and surface usage metrics (top queries, click-through rate, unanswered questions) so we can keep improving both the intranet content and the bot itself.
Technical outline
– SharePoint Online (modern sites).
– Preferred stack: Azure OpenAI / OpenAI GPT-4 (or similar LLM) with a vector store such as Azure Cognitive Search, Pinecone or Redis.
– Authentication should respect Microsoft 365 permissions; users may only receive information they already have access to.
– Front-end can be a SharePoint Framework (SPFx) web part or an adaptive card inside Viva Connections; as long as the UI is clean and blends with the existing intranet branding, I’m flexible.
– All code will be stored in our private Git repository with clear setup instructions for dev, test and prod.
Deliverables
1. Architecture diagram and stack recommendation.
2. Working chatbot deployed in our test tenant.
3. Documentation covering setup, permissions, prompts, and how to add new content sources.
4. A short video walkthrough plus live hand-over session.
Acceptance criteria will be a demo where I ask the bot to find a current policy, receive an answer citing the correct page, and view metrics recorded in the analytics dashboard.
If you have shipped a similar SharePoint or Microsoft 365 conversational solution before, I’d love to see it when we talk.
Core interactions
• Content discovery – when someone asks a question, the bot should crawl or query published intranet pages, list documents, and any connected knowledge bases, then return the most relevant pages, files, or list items.
• User support – conversational answers must be generated in natural language, summarising or explaining information from the source page so users know immediately why that item was recommended.
• Data analytics – capture and surface usage metrics (top queries, click-through rate, unanswered questions) so we can keep improving both the intranet content and the bot itself.
Technical outline
– SharePoint Online (modern sites).
– Preferred stack: Azure OpenAI / OpenAI GPT-4 (or similar LLM) with a vector store such as Azure Cognitive Search, Pinecone or Redis.
– Authentication should respect Microsoft 365 permissions; users may only receive information they already have access to.
– Front-end can be a SharePoint Framework (SPFx) web part or an adaptive card inside Viva Connections; as long as the UI is clean and blends with the existing intranet branding, I’m flexible.
– All code will be stored in our private Git repository with clear setup instructions for dev, test and prod.
Deliverables
1. Architecture diagram and stack recommendation.
2. Working chatbot deployed in our test tenant.
3. Documentation covering setup, permissions, prompts, and how to add new content sources.
4. A short video walkthrough plus live hand-over session.
Acceptance criteria will be a demo where I ask the bot to find a current policy, receive an answer citing the correct page, and view metrics recorded in the analytics dashboard.
If you have shipped a similar SharePoint or Microsoft 365 conversational solution before, I’d love to see it when we talk.
Related categories:
C# Programming
Sharepoint
Azure
Node.js
Redis
Data Analytics
Microsoft 365
Azure OpenAI