Workflow Automation · Business Operations · Systems Integration
Building an Automated Workflow Assessment and Lead Management System
The Workflow Assessment on this website: a five-step discovery tool that turns a description of manual work into a conceptual Automation Blueprint, a structured lead and a follow-up workflow.
- Industry
- Software engineering services
- Project type
- Assessment, lead capture and operations tooling
- Attribution
- PathWebSolutions Internal Project
- Status
- Live on this website
Simplified system overview
- Visitor
- Assessment → Blueprint
- SupabaseStorage & queues
- EmailExternal serviceAdmin reviewPeople
01 / Overview
Project overview
PathWebSolutions built an interactive Workflow Assessment that helps visitors describe manual business processes, understand potential automation opportunities and generate a conceptual Automation Blueprint.
The platform connects discovery, lead capture, email notifications, scheduling and internal lead management.
PathWebSolutions Internal Project. Built and operated by PathWebSolutions. You can use it on this website.
02 / Challenge
The challenge
Traditional contact forms provide limited technical context. A business may request automation without clearly describing:
- The inputs that start the work
- The manual tasks involved
- The systems the results go to
- How much work is processed
- Where human approval is required
- How exceptions are handled
- Integration constraints
Each missing detail means another round of follow-up questions before technical discovery can begin.
03 / Solution
The solution
An interactive, structured assessment replaces the open-ended form. Visitors describe the work one step at a time, see their workflow drawn as they go and receive an Automation Blueprint at the end. The same structured answers become the lead the team reviews.
04 / Capabilities
Key capabilities
Five-step assessment
Problem, workflow, workload, controls and outcome, one focused step at a time.
Live workflow visualization
Inputs, manual steps and destinations are drawn as the visitor chooses them.
Deterministic automation fit
The same answers always produce the same classification: strong, partial, or technical discovery recommended.
Manual-workload calculations
Volume and handling time become monthly and annual manual-hour estimates.
Suggested automation architecture
The Blueprint outlines how the workflow could be automated.
Human-review recommendations
It identifies where people should stay involved.
Lead capture
Contact details are requested only after the Blueprint is shown.
Supabase persistence
The assessment and the lead are saved together in PostgreSQL.
Transactional email via Resend
The visitor receives their Blueprint, and the team receives a notification.
Booking integration
Visitors can book a review call right after submitting.
Private lead dashboard
An authenticated operations dashboard to review leads, update their status and keep internal notes.
First-party conversion analytics
Anonymous measurement of the assessment funnel, without third-party trackers.
05 / Architecture
Technical architecture
How the parts connect. The animated example follows real stages of the system; every step is also described in text.
- Visitor
- Workflow Assessment5 steps
- Validation & rules engine
- Automation Blueprint
- Lead submissionTurnstile protected
- Next.js serverValidates & recalculates
- Supabase PostgreSQLDatabase
- ResendTransactional emailAdmin dashboardAuthenticated
- Customer emailLead reviewPeople
- Follow-up
- Data flow
- Processing
- Outcome
- Storage & queues
- External service
- People
How it works
Describe the workflow
The visitor works through five short steps while a live diagram shows the inputs, manual steps and destinations they choose.
Evaluate with rules
Validation and a deterministic rules engine produce the automation fit, workload estimate and recommendations.
Share the Blueprint first
The visitor sees the full Automation Blueprint before being asked for contact details.
Submit safely
Cloudflare Turnstile is verified, and the Next.js server validates every answer and recalculates the Blueprint, ignoring values from the browser.
Store, notify and follow up
The assessment and lead are saved in Supabase, Resend delivers the visitor and team emails, and administrators review the lead in the private dashboard.
06 / Decisions
Engineering decisions
Deterministic scoring instead of unnecessary LLM calls
Classification, workload and recommendations come from explicit rules. Results are instant, consistent and explainable, with no per-request model cost or model output to check.
Server-side validation and recalculation
Browser calculations power the live experience only. The server validates every answer and recomputes the Blueprint, so a modified request cannot change a lead’s classification.
Database persistence with restricted access
Server-only code saves the assessment and lead together. Row-level security is enabled with no public policies, so the public key can neither read nor write them.
Turnstile and submission protection
Cloudflare Turnstile is checked on the server before anything is saved or sent, and a submission ID ensures that retries never create duplicate leads or emails.
Public and private features kept apart
The public assessment and the operations dashboard share data, not access. The dashboard requires Supabase Auth plus a server-side allowlist and is never indexed.
First-party, privacy-conscious analytics
Funnel events are anonymous, first-party and stored in the same database. There are no third-party trackers or personal data in events, and visitors can opt out.
07 / Stack
Technologies
08 / Outcomes
Business outcomes
This is an internal implementation; no conversion improvements are claimed.
- General inquiries become structured requirements.
- Each completed assessment produces a conceptual automation blueprint.
- Leads are stored and organized in one place.
- Confirmation and notification emails are sent automatically.
- An internal dashboard supports review and follow-up.
- Assessment funnel activity can be measured.
09 / Takeaways
What this demonstrates
- Turning a discovery conversation into a structured product
- Rules-based evaluation where AI is not needed
- A secure path from the browser to the database and the inbox
- Building the operations tooling behind a public feature
Discuss a similar project
Have a similar engineering challenge?
Tell us about the workflow, integration, or technical problem you’re trying to solve.