Case Study: MPI Systems
Case Study: MPI Systems
How an 85-Person Manufacturer Cut Information Search Time in Half to Increase Employee Productivity
This case study is derived from a technical paper presented by
Aaron Phipps (President @ MPI) and Patrick Byrne (CEO @ Annora) at the 2025 Investment Casting Institute Technical Conference
How an 85-Person Manufacturer Cut Information Search Time in Half to Increase Employee Productivity
This case study is derived from a technical paper presented by
Aaron Phipps (President @ MPI) and Patrick Byrne (CEO @ Annora) at the 2025 Investment Casting Institute Technical Conference
The Company
The Company
MPI Systems is a precision equipment manufacturer with 85 employees, specializing in high-quality custom solutions for the wax room. With over 50 years of equipment in the field, their team supports a vast installed base while continuing to engineer and manufacture new products.
Like many manufacturers, MPI faced a growing challenge: critical knowledge was scattered across multiple systems and increasingly concentrated in the heads of experienced employees approaching retirement.
MPI Systems is a precision equipment manufacturer with 85 employees, specializing in high-quality custom solutions for the wax room. With over 50 years of equipment in the field, their team supports a vast installed base while continuing to engineer and manufacture new products.
Like many manufacturers, MPI faced a growing challenge: critical knowledge was scattered across multiple systems and increasingly concentrated in the heads of experienced employees approaching retirement.
The Challenge
The Challenge
MPI Systems was experiencing what most manufacturers experience but rarely quantify:
Information was everywhere but employees couldn't find it.
Technical documentation lived in SharePoint. SOPs were in GembaDocs. Quotes and orders were buried in local file servers and the ERP. Past solutions to customer problems existed only in email threads or the memories of senior staff.
The result:
Employees spent nearly 2 hours per day searching for information across disconnected systems
Senior employees were constantly interrupted to answer questions that should have been easy to find
New hires took months to become productive because institutional knowledge wasn't accessible
Customer response times suffered when answers required hunting through multiple systems
With 25% of the manufacturing workforce nationally aged 55 or older, MPI knew they needed to capture and systematize knowledge before it walked out the door.
MPI Systems was experiencing what most manufacturers experience but rarely quantify:
Information was everywhere but employees couldn't find it.
Technical documentation lived in SharePoint. SOPs were in GembaDocs. Quotes and orders were buried in local file servers and the ERP. Past solutions to customer problems existed only in email threads or the memories of senior staff.
The result:
Employees spent nearly 2 hours per day searching for information across disconnected systems
Senior employees were constantly interrupted to answer questions that should have been easy to find
New hires took months to become productive because institutional knowledge wasn't accessible
Customer response times suffered when answers required hunting through multiple systems
With 25% of the manufacturing workforce nationally aged 55 or older, MPI knew they needed to capture and systematize knowledge before it walked out the door.
The Solution
The Solution
Annora AI partnered with MPI Systems to build an AI-powered knowledge retrieval system that connects to their existing data sources—without replacing any systems or requiring data migration.
How it works:
The system uses hybrid search technology that combines traditional keyword matching with semantic AI search. This means employees can search using exact part numbers OR natural language questions like "What was the fix for the cooling issue on the 2019 Model X?"
Data sources connected:
SharePoint (technical docs, engineering drawings)
GembaDocs (SOPs, quality documentation)
Local file servers (ERP-generated quotes, orders, POs)
ERP system (live order status, project information)
Recorded knowledge sessions from retiring employees
Key feature: Every answer cites its sources with direct links to the original documents. Users can verify information and access full context in seconds.
Annora AI partnered with MPI Systems to build an AI-powered knowledge retrieval system that connects to their existing data sources—without replacing any systems or requiring data migration.
How it works:
The system uses hybrid search technology that combines traditional keyword matching with semantic AI search. This means employees can search using exact part numbers OR natural language questions like "What was the fix for the cooling issue on the 2019 Model X?"
Data sources connected:
SharePoint (technical docs, engineering drawings)
GembaDocs (SOPs, quality documentation)
Local file servers (ERP-generated quotes, orders, POs)
ERP system (live order status, project information)
Recorded knowledge sessions from retiring employees
Key feature: Every answer cites its sources with direct links to the original documents. Users can verify information and access full context in seconds.
Real-World Use Cases
Real-World Use Cases
1. Faster Quoting: Sales and estimating teams search previous quotes and orders to build new quotes faster. Instead of recreating pricing from scratch or hunting for similar past projects, they find relevant history in seconds.
2. Better Customer Service: When a customer calls about an order or a piece of equipment, service reps instantly pull up the full history—original quote, order details, any past service issues—without putting customers on hold while they search.
3. Shop Floor Access to SOPs: Manufacturing personnel search for relevant procedures and specifications while on the floor, without walking back to an office or interrupting a supervisor.
1. Faster Quoting: Sales and estimating teams search previous quotes and orders to build new quotes faster. Instead of recreating pricing from scratch or hunting for similar past projects, they find relevant history in seconds.
2. Better Customer Service: When a customer calls about an order or a piece of equipment, service reps instantly pull up the full history—original quote, order details, any past service issues—without putting customers on hold while they search.
3. Shop Floor Access to SOPs: Manufacturing personnel search for relevant procedures and specifications while on the floor, without walking back to an office or interrupting a supervisor.
The Results
The Results
Annora and MPI conducted a controlled study with 30 employees across departments. Each employee completed a 10-question assessment requiring information from multiple systems—first using traditional search methods, then after a 30-minute training session on the AI system.
At full adoption across the organization:
3-4 FTE equivalent in productivity gains (3.5-4.7% improvement across 85 employees)
250%+ ROI in the first year of deployment
30 minutes of training required for employees to use the system effectively
New hires saw up to 79% improvement in finding correct information
Annora and MPI conducted a controlled study with 30 employees across departments. Each employee completed a 10-question assessment requiring information from multiple systems—first using traditional search methods, then after a 30-minute training session on the AI system.
At full adoption across the organization:
3-4 FTE equivalent in productivity gains (3.5-4.7% improvement across 85 employees)
250%+ ROI in the first year of deployment
30 minutes of training required for employees to use the system effectively
New hires saw up to 79% improvement in finding correct information
Built Into Every Workflow
Built Into Every Workflow
This knowledge retrieval capability isn't a standalone product—it's built into every system Annora deploys.
Whether your team is using our quoting system, inventory management, purchasing, or scheduling tools, relevant information is surfaced where they're already doing the work. No context switching. No hunting across systems.
The knowledge follows the workflow.
This knowledge retrieval capability isn't a standalone product—it's built into every system Annora deploys.
Whether your team is using our quoting system, inventory management, purchasing, or scheduling tools, relevant information is surfaced where they're already doing the work. No context switching. No hunting across systems.
The knowledge follows the workflow.
Published Research Paper
Published Research Paper
This case study is based on a technical paper co-authored by Aaron Phipps (President, MPI Systems), Patrick Byrne (CEO, Annora AI), and Dr. Wes Teskey (CTO, Annora AI).
"Bridging the Knowledge Gap: AI-Powered Information Systems in Manufacturing" was presented at the Investment Casting Institute's 72nd Technical Conference & Expo in September 2025.
Read the full technical paper →
This case study is based on a technical paper co-authored by Aaron Phipps (President, MPI Systems), Patrick Byrne (CEO, Annora AI), and Dr. Wes Teskey (CTO, Annora AI).
"Bridging the Knowledge Gap: AI-Powered Information Systems in Manufacturing" was presented at the Investment Casting Institute's 72nd Technical Conference & Expo in September 2025.
Read the full technical paper →
Ready to See It in Action?
Ready to See It in Action?
See how an AI-powered knowledge system could work for your manufacturing operation.
We'll show you:
How the system connects to your existing data sources
A live demonstration of hybrid search in action
How knowledge retrieval integrates with quoting, customer service, and shop floor workflows
ROI projections based on your team size
See how an AI-powered knowledge system could work for your manufacturing operation.
We'll show you:
How the system connects to your existing data sources
A live demonstration of hybrid search in action
How knowledge retrieval integrates with quoting, customer service, and shop floor workflows
ROI projections based on your team size
© 2024 Annora. All rights reserved.
© 2024 Annora. All rights reserved.
© 2024 Annora. All rights reserved.