Sample Builds
Three self-initiated demos showing how automation can solve common operational problems.
These are self-initiated demo builds, not completed client engagements.
We built each of these ourselves to demonstrate how we'd solve a real, common operational problem for this kind of business. No named company was involved, and the results below are projected based on realistic assumptions about the manual process being replaced, not measured client outcomes.
Automated Shipment Tracking & Client Notification System
Scenario: 35-person freight forwarding company, US-Mexico corridor
Industry: Logistics / Freight Forwarding
Tool: n8n (workflow automation)
Type: Proof-of-Concept Build
The Business
A 35-person freight forwarding company specialises in cross-border logistics along the US–Mexico corridor. At any given time, the team manages 40–60 active shipments across three carrier partners, each delivering status updates in a different format: plain-text email, PDF attachment, and Excel spreadsheet.
The Problem
Every working day, a junior operations staff member manually processed incoming carrier emails:
- Open each carrier email or attachment (plain text, PDF, or Excel)
- Read and interpret the shipment status, location, and ETA
- Locate the matching row in the master Google Sheets tracker
- Update the status, ETA, location, and delay notes manually
- If the shipment was delayed, draft and send a client notification email
Pain points
| Area | Issue | Impact |
|---|---|---|
| Time cost | 3 hrs/day of manual processing | 15 hrs/week absorbed into low-value admin |
| Error rate | Manual copy-paste between systems | Wrong clients notified; ETAs not updated |
| Inconsistency | 3 carriers, 3 different formats | No standardised intake process |
| Delayed alerts | Client emails written and sent manually | Clients sometimes notified hours late |
The Solution
An automated n8n workflow monitors a Gmail inbox, intelligently parses each carrier's unique format, updates the Google Sheets tracker, and sends client delay notifications automatically without any human involvement.
How it works
Key technical components
- •n8n self-hosted on VPS
- •Gmail API via OAuth2
- •pdf.co API for PDF extraction
- •Google Sheets API
- •Switch + IF nodes for conditional routing
- •JavaScript Code nodes for regex extraction
Projected Impact
| Metric | Result |
|---|---|
| Ops time eliminated | ~15 hrs/week |
| Labour cost saved | ~$1,500/mo |
| Delay notification time | <5 minutes |
| Manual data entry touchpoints | 0 |
| Carrier formats handled | 3 (text, PDF, Excel) |
| Duplicate notification prevention | 100% |
Automated Client Intake Processing System
Scenario: 25-person legal services firm, UK-based SME client work
Industry: Legal / Professional Services
Tool: Django (Python)
Type: Proof-of-Concept Build
The Business
A 25-person legal services firm specialising in contract review, client intake, and compliance documentation for SME clients onboards 15–20 new clients per month. Each requires a structured intake process managed by the paralegal team.
The Problem
Every time a new client signed up, a paralegal manually worked through these steps:
- Open the client intake PDF submitted by the prospective client
- Manually read and extract key data fields: name, email, matter type, referral source
- Create a new client folder in Google Drive with a consistent naming convention
- Copy all client details into the master Google Sheets client log
- Draft a welcome email and engagement letter from a Word template
- Send both documents to the supervising partner for review
Pain points
| Area | Issue | Impact |
|---|---|---|
| Time cost | 45 mins per new client intake | ~15 hrs/month absorbed into pure admin |
| Error rate | Manual copy-paste between PDF and Sheet | Wrong names in letters, missing fields |
| Inconsistency | Folder naming done by hand each time | Inconsistent Drive structure across clients |
| Bottleneck | Partner review delayed by slow prep | New client response times stretched to days |
The Solution
A custom Django application watches a designated Google Drive folder for new intake PDFs, processes each one end-to-end without human intervention, and notifies the partner the moment a new client is ready for review.
How it works
Key technical components
- •Django (Python): core application framework
- •pdfplumber: PDF text extraction
- •google-api-python-client: Drive folder creation and Sheets logging
- •Gmail SMTP: transactional email delivery
- •Django Management Command: scheduled polling loop via cron
Projected Impact
| Metric | Result |
|---|---|
| Paralegal time eliminated | ~15 hrs/month |
| End-to-end intake processing | <2 minutes |
| Labour cost saved | £525/month |
| Manual data entry steps | 0 |
| Duplicate email prevention | 100% |
| Test intake forms processed | 5 |
AI-Powered CV Processing & Candidate Scoring System
Scenario: 40-person staffing agency, GCC logistics/operations recruitment
Industry: Staffing & Recruitment
Tool: Django + OpenAI GPT-4o-mini
Type: Proof-of-Concept Build
The Business
A 40-person staffing agency specialises in placing candidates in logistics and operations roles across the GCC region. At peak hiring periods, the team receives 80–120 CVs per week across five platforms, all funnelling into a single shared Gmail inbox.
The Problem
For every CV received, a junior recruiter manually:
- Opened the CV email and downloaded the PDF attachment
- Read the CV and extracted key fields: name, contact details, experience, skills, location
- Copied all data into the Google Sheets candidate tracker
- Manually scored the candidate as Hot, Warm, or Cold based on gut feel
- Drafted and sent a templated acknowledgement email to the candidate
Pain points
| Area | Issue | Impact |
|---|---|---|
| Volume | 80–120 CVs/week across 5 platforms | Recruiters spending 20+ hrs/week on admin alone |
| Inconsistency | Manual scoring based on gut feel | Qualified candidates scored differently by each recruiter |
| Speed | Response times stretched to 3–4 days | Strong candidates lost to faster-responding agencies |
| Data quality | Copy-paste from PDF to Sheet | Missing fields, formatting errors, incomplete records |
| Bottleneck | Best candidates buried under volume | No prioritisation: Hot leads treated same as Cold |
The Solution
A Django application with an integrated OpenAI scoring engine monitors the Gmail inbox, processes each CV end-to-end, and delivers a scored, logged, and acknowledged candidate record without any recruiter involvement.
How it works
Key technical components
- •Django (Python): core application framework and admin
- •OpenAI GPT-4o-mini: AI candidate scoring (Hot / Warm / Cold with reasoning)
- •pdfplumber: PDF text extraction
- •Gmail API: inbox monitoring and attachment download
- •google-api-python-client: Google Sheets logging
- •Gmail SMTP: acknowledgement email delivery
Projected Impact
| Metric | Result |
|---|---|
| Recruiter admin time eliminated | ~20 hrs/week |
| CV processing time | <30 seconds |
| Scoring consistency | 100% (AI criteria, not gut feel) |
| Manual data entry steps | 0 |
| Candidate acknowledgement | Instant |
| Prioritisation | Hot candidates surfaced automatically |