Skip to main content
Insights hub

Case study · Data, AI & ML

AI Knowledge Assistant for a Singapore Professional Services Firm

How a 40-Person Firm Reduced Internal Knowledge Search Time by 65% and Accelerated Proposal Creation by More Than 60%

Professional Services

Illustrative image of professional-services consultants reviewing internal knowledge and proposal documents

Executive Summary

A growing professional-services firm was spending valuable billable time searching for information scattered across SharePoint sites, shared drives, templates, project folders, and legacy proposal documents.

ExpertsPro designed an AI-powered Knowledge Assistant integrated with Microsoft Teams to help employees quickly find trusted information, access approved content, and generate proposal outlines from existing knowledge assets while maintaining governance, security, and human oversight.

Estimated Business Impact

  • Approximately 13 staff hours saved per day
  • Approximately 3,000 staff hours returned annually
  • Proposal first-draft effort reduced from 4 hours to approximately 1.5 hours
  • Faster onboarding and improved knowledge transfer
  • Reduced dependence on senior staff for repetitive questions
  • Better reuse of approved content and organizational knowledge

Client Profile

CategoryDetails
IndustryProfessional Services (Consulting, Engineering, Architecture, Marketing, IT Services)
RegionSingapore (also applicable to Australia, UK, and USA)
Company SizeApproximately 40 Employees (ideal for 20–150 employees)
Engagement ModelPilot followed by Managed AI Operations
Delivery Timeline5 Weeks
TeamAI Engineer, Data Engineer, Delivery Lead
Estimated InvestmentSGD 15,000–25,000 Pilot + Ongoing Support
TechnologyMicrosoft 365, SharePoint, Microsoft Teams, Azure OpenAI, Permission-Aware Knowledge Search (RAG), Python

Business Challenge

Over time, the firm accumulated years of valuable knowledge across:

  • Project documentation
  • Proposal libraries
  • Methodologies
  • Templates
  • Internal guides
  • Policies and procedures
  • Best-practice documents

Although the information existed, employees often struggled to find the right document or example when needed.

With approximately 40 employees spending around 30 minutes each day searching for information, the organization was losing nearly 20 staff hours every day on knowledge retrieval alone.

Key Business Problems

  • Time lost searching for templates, methodologies, and past work
  • Senior staff repeatedly answering the same internal questions
  • Proposal drafting frequently starting from a blank page
  • New employees taking longer to become productive
  • Knowledge spread across multiple repositories and file structures
  • Inconsistency in using approved organizational content

Solution Overview

ExpertsPro implemented an AI Knowledge Assistant embedded directly inside Microsoft Teams.

The assistant enables employees to ask natural-language questions and receive answers drawn exclusively from approved internal content while respecting existing permissions.

In addition to answering questions, the assistant can generate proposal outlines using previously approved proposal components and organizational knowledge.

Every response includes source references, allowing staff to verify and trust the information being provided.

See the Solution in Action

See how the AI Knowledge Assistant brings approved internal knowledge into Microsoft Teams, supports proposal drafting, and keeps source references, access permissions, and employee review at the centre of the workflow.

AI knowledge assistant case study opening screen showing information scattered across internal repositories

Example Questions Employees Can Ask

  • Show me our latest manufacturing proposal.
  • What is our standard cloud migration methodology?
  • Find examples of cybersecurity assessments completed in the past two years.
  • What are our travel and expense reimbursement guidelines?
  • Create a proposal outline for a retail modernization engagement.
  • Where can I find our project handover checklist?
  • What are the standard deliverables for a data analytics engagement?

How the Solution Works

1. Define Scope

The organization selects which repositories should be included:

  • SharePoint libraries
  • Approved shared drives
  • Methodology documents
  • Templates
  • Policies
  • Standard operating procedures
  • Proposal content libraries

Personal files and individual email content remain excluded.

2. Prepare Content

Proposal sections and reusable content are reviewed and sanitized where necessary.

Confidential client names and proprietary customer information are removed before inclusion in the knowledge library.

3. Build Searchable Knowledge

Documents are indexed using AI-powered semantic search.

Instead of searching only for keywords, employees can search using business questions and natural language.

4. Enforce Existing Permissions

The system respects existing Microsoft 365 permissions.

Employees can only access information they are already authorized to view. The AI assistant does not grant access to restricted content.

5. Answer Questions and Draft Content

Through Microsoft Teams, employees can:

  • Ask questions
  • Locate documents
  • Summarize procedures
  • Retrieve examples
  • Generate proposal outlines
  • Reuse approved content

Responses include links to the original source documents for verification.

6. Human Review

AI-generated outputs are always reviewed by employees before being shared internally or externally.

The solution assists decision-making and content preparation but does not replace human judgment.

Solution Workflow

Illustration of a Teams question flowing through approved documents and permission-aware search to a cited answer or proposal for human review

Employee Question in Teams → Permission-Aware Knowledge Search → Answer with Source Citations OR Proposal Outline Generation → Employee Review → Approved Internal or Client Use

Human Oversight & Governance

AI Handles

  • Knowledge search
  • Document retrieval
  • Summaries
  • Source citations
  • Proposal outline creation
  • Knowledge discovery

Employees Handle

  • Scope definition
  • Content approval
  • Proposal customization
  • Client-specific decisions
  • Final review and approval

Data Protection

All data remains within the organization's Microsoft 365 and Azure environment.

No customer documents are used to train Azure OpenAI models.

The organization maintains ownership and control of all content and access permissions.

Delivery Approach

Week 1

  • Discovery workshops
  • Content assessment
  • Scope definition
  • Selection of approved repositories

Weeks 2–3

  • Document preparation
  • Knowledge indexing
  • Semantic search testing
  • Validation against approximately 200 sample questions

Week 4

  • Microsoft Teams integration
  • Proposal drafting capabilities
  • User testing and refinements

Week 5

  • Pilot deployment
  • User training
  • Feedback collection
  • Handover and production readiness

Estimated Business Outcomes

Knowledge Search Efficiency

MeasureBeforeAfter
Knowledge search / staff hours per day~20~7
Proposal first-draft effort~4 hours~1.5 hours

Capacity Returned

Approximately 13 Staff Hours Saved Daily

Knowledge Accessibility

  • Employees find information faster
  • New employees become productive sooner
  • Senior staff spend less time answering repetitive questions

Content Consistency

  • Greater use of approved methodologies
  • Improved proposal quality and consistency
  • Reduced duplication of effort

Annual Capacity Impact

13 Hours Saved Daily × ~230 Working Days ≈ 3,000 Staff Hours Per Year

This is equivalent to recovering nearly 1.5 full-time employees' worth of productive capacity without increasing headcount.

Those hours can be redirected toward:

  • Client delivery
  • Revenue-generating work
  • Business development
  • Innovation initiatives

Why AI Was a Good Fit

This process was an ideal candidate for AI because it is:

  • High volume
  • Repetitive
  • Knowledge intensive
  • Document driven
  • Easy to measure
  • Dependent on information that already exists

The challenge was not a lack of knowledge. The challenge was finding and reusing knowledge efficiently.

Best-Fit Organizations

This solution is particularly valuable for:

  • Consulting firms
  • Engineering organizations
  • Architecture and design firms
  • IT services companies
  • Legal and advisory practices
  • Marketing agencies
  • Professional-services organizations using Microsoft 365

Especially where:

  • Knowledge is distributed across repositories
  • Employees repeatedly search for information
  • Proposal development consumes significant time
  • New-employee onboarding relies heavily on senior employees

Why ExpertsPro

ExpertsPro combines:

  • AI & Machine Learning Engineering
  • Data & Analytics Expertise
  • Microsoft 365 Solutions
  • Azure AI Services
  • Knowledge Management Modernization
  • Managed AI Operations

to help organizations transform existing knowledge assets into measurable business value while maintaining governance, security, and operational control.

Ready to Unlock Knowledge Already Hidden Inside Your Organization?

If your team spends hours searching for information, recreating proposals, or answering the same internal questions repeatedly, ExpertsPro can help identify high-impact AI opportunities.