AI Automation
Proposal
A low-risk, fixed-fee discovery and three focused automation tracks that remove manual, repetitive, error-prone work across customer service and product design — built to the security posture your aerospace and defense customers require.
Phil Therien — Partner and Chief Revenue Officer, Webisoft Technologie Inc.
Context and intent
Overview
Lynch Fluid Controls designs and manufactures hydraulic motion control manifolds and integrated hydraulic systems, running the business on Microsoft Business Central with Copilot in some workflows and Power BI for reporting. In our intro call we looked at where AI and automation could remove manual, repetitive, error-prone work across customer service and product design, while fitting the security posture your aerospace and defense customers require.
This proposal outlines a low-risk discovery phase and the automation tracks we would evaluate together. The intent is to prove value on a small, measurable pilot before any broader rollout, and to build on your early Copilot use rather than replace how your team works.
Discovery Fee
$2,500CAD — fixed
Discovery Length
2–3weeks
Automation Tracks
3scoped together
Per-Track Build
<$10KCAD quick-win scope
The discovery fee is credited in full toward the first implementation track if you proceed. There is no obligation to continue past discovery.
The approach
Prove value on a small, measurable pilot before any broader rollout — and build on your early Copilot use rather than replace how your team works.
Phil Therien
Partner and Chief Revenue Officer, Webisoft
How we work
Guiding Principles
Start small and prove value
Quick-win pilots with measurable time savings come before any company-wide deployment.
Fit the existing stack
Solutions integrate with Business Central and your design management system through their APIs rather than replacing them.
Keep a human in the loop
AI drafts, suggests, and looks up; your team reviews and decides — especially on the design side.
Security first
Architecture designed for data residency, access control, logging, and least-privilege access to meet CMMC and CPCSC requirements. Closed-box models on controlled infrastructure are the default for sensitive data.
Targeted over general
Custom, domain-specific tools built for your data flows deliver higher accuracy and ROI than general-purpose assistants.
$2,500 CAD fixed fee — credited toward the first track
Discovery Phase
A fixed-fee discovery of $2,500 CAD, credited in full toward the first implementation track if you proceed. Over 2 to 3 weeks we run workshops with your customer service, engineering, and IT stakeholders and take scoped access to the relevant systems. You finish discovery with a clear scope, a plan, and no obligation to continue.
What we do
Stakeholder workshops
Sessions with customer service, engineering, and IT stakeholders.
Scoped system access
Controlled access to Business Central and the design management system — only what discovery needs.
Feasibility vs. security review
Validate the highest-value automation targets against your CMMC / CPCSC-driven security requirements.
What you get
Confirmed automation targets
The highest-value tracks, sequenced by impact and feasibility.
ACAgreed with your stakeholders before any build starts.
Success metrics
The measures every pilot is judged against before go-live.
A clear scope and plan
Priced, sequenced, and yours to act on — with no obligation to continue.
Three focused automations, sequenced your way
Opportunity Tracks
Order Entry & Customer Inquiry Handling
Microsoft Business CentralToday
Staff manually research price checks, stock availability, and order status inside Business Central, and transcribe purchase order requests by hand. It is slow and prone to errors from typos on similar part numbers.
What we build
An AI first-line support agent that reads incoming inquiries, pulls live data from Business Central, and either responds directly or escalates based on request type. Communication actions are logged back into your systems automatically. Before go-live, the agent runs in a closed loop alongside your team through a trial dashboard, so you compare its output to human performance side by side before trusting it with customers.
Expected impact
Faster replies, fewer manual touchpoints per order and inquiry, and a real reduction in entry errors.
Product Design & Validation Support
Design management systemToday
Engineers spend hours designing hydraulic manifolds that often resemble designs already held in your design management system. At an engineering cost of $80 to $100 per hour, saving even 10 hours per design is meaningful — and it helps preserve design knowledge as experienced staff retire.
What we build
A recommendation engine that scans your existing design library for feature-based similarities and application tags such as agricultural or mining, then notifies engineers of matching or near-duplicate designs so they reuse rather than rebuild. It runs as a daily scan and stays a suggestion tool: engineers review and adapt every recommendation. Metadata such as material, size, features, and application tags guides the search, and part of this track is making that tagging effective.
Expected impact
Hours saved per design at $80–$100 / hour engineering cost, and design knowledge preserved as experienced staff retire.
AI Ramp & Fit Mapping
Copilot + your data flowsToday
You already use Copilot in places, but there is no map of where general-purpose tools serve you well versus where custom, API-integrated solutions would deliver more.
What we build
A mapping of where general-purpose tools fit and where custom solutions built for your specific data flows deliver higher accuracy and ROI. The output is a targeted adoption plan rather than broad, unfocused tooling — a practical path for your team to ramp onto AI in a way that fits the business.
Expected impact
A targeted adoption plan and a practical AI ramp for the team, instead of broad, unfocused tooling.
Tracks at a glance
| Criterion | T1 — Orders & Inquiries | T2 — Design Support | T3 — AI Ramp |
|---|---|---|---|
| Integrates with | Business Central | Design mgmt system | Copilot + stack |
| Human stays in the loop | Yes | Yes | Yes |
| Closed-loop trial before go-live | Yes | Yes | No |
| Primary payoff | Fewer entry errors | Hours saved per design | Focused adoption |
| Quick-win build under $10K | Yes | Yes | Yes |
CMMC and CPCSC by design
Security & Compliance
Given your aerospace and defense customers, any solution touching sensitive email or system data is built around data residency, role-based access control, full logging of communication actions, and least-privilege access, consistent with CMMC and CPCSC standards. A closed-box model running on controlled infrastructure is our default for sensitive workloads.
Built in from the start
- Data residency by design
- Role-based access control
- Full logging of communication actions
- Least-privilege access
- Closed-box models on controlled infrastructure
- Onshore in Canada with a senior in-house team
These requirements may add to project cost, but we plan for them from the start rather than treating them as blockers. The specific architecture is confirmed with your IT team during discovery.
Prove, then scale
Phased Roadmap
2–3 weeks
Discovery
Workshops, scoped system access, target selection, security design, and agreed success metrics.
Per track
Pilot build
Implementation of the first agreed track, integrated to Business Central or your design management system.
Before go-live
Closed-loop trial
The AI runs alongside your team with a comparison dashboard — its output measured against human performance before it touches a customer.
On proven value
Go-live & iterate
Deploy, measure against the agreed success metrics, then evaluate the next track.
All figures in CAD
Pricing
| Item | Effort | Amount | |
|---|---|---|---|
| Discovery phaseFixed fee. Credited in full toward the first implementation track if you proceed. | Fixed fee. Credited in full toward the first implementation track if you proceed. | 2–3 wks | $2,500 |
| Implementation, per trackInitial quick-win scope, integrated to your systems, with closed-loop trial. | Initial quick-win scope, integrated to your systems, with closed-loop trial. | per track | under $10,000 |
| Security-hardened architectureFor sensitive workloads. Scoped and quoted during discovery, before you commit. | For sensitive workloads. Scoped and quoted during discovery, before you commit. | as scoped | quoted in discovery |
All figures in CAD. Each track stands on its own, so you can sequence them in whatever order makes the most sense for the business.
Validate value at every step
Engagement Model
Inclusions
04- Fixed-fee discovery with workshops and scoped system access
- Per-track implementations you approve one at a time
- Closed-loop trial with a comparison dashboard before go-live
- Discovery fee credited in full toward the first track
Exclusions
03- No obligation to proceed past discovery
- No company-wide deployment before a measured pilot
- No replacement of Business Central or your design system
Assumptions
03- Scoped access to relevant systems during discovery
- Customer service, engineering, and IT stakeholders available for workshops
- Security architecture confirmed with your IT team during discovery
Fixed-fee discovery to start, followed by per-track implementations so you validate value at each step before committing to the next.
Anticipated questions
Next Steps
Three steps between this document and a discovery kickoff.
NDA from Lynch
Formalize discussions and enable data sharing.
Follow-up call — July 17, 10:00 AM
Review this proposal with your internal stakeholders and AI champions.
Discovery kickoff
On your go-ahead, we schedule the kickoff and confirm workshop participants.
Phil Therien, Partner and Chief Revenue Officer — hello@webisoft.com
Webisoft · Montreal, QC · webisoft.com
Confidential — Lynch Fluid Controls · AI Automation Proposal