Client details anonymised at their request. References available — we're happy to connect you with existing clients in your sector.
A one-man plumbing business was spending money on Google Ads and Checkatrade to get calls — then missing them while on jobs. Evenings were spent returning calls to people who had already booked someone else. An estimate suggested 6–8 leads lost per month at an average job value of £280.
AI call handler via Bland.ai: answers every call 24/7, qualifies the enquiry (job type, urgency, location), confirms a callback time, and sends a WhatsApp summary to the client's phone. Integrated with Google Calendar for availability context. No new apps — just a forwarded number.
A Google Sheets-based customer log captures job dates and service types at completion. An n8n workflow triggers personalised WhatsApp and email reminders at 11 months (boiler service due), 10.5 months (gas certificate renewal), and a follow-up 7 days later if no response. Each message is personalised with the customer's name, property address, and last service date.
An electrical contractor with a team of field engineers was spending every Friday afternoon chasing timesheets by phone and WhatsApp, manually entering hours into a spreadsheet, and reconciling discrepancies before payroll. The office manager estimated 3–4 hours lost every week, with payroll errors creating additional rework.
WhatsApp-triggered n8n workflow: engineers message their hours in any format — the Claude API parses the message and extracts structured data. Automatically logged to a Google Sheet. Non-responders receive a WhatsApp chase at Friday midday. A payroll-ready summary emails to the office manager by 4pm without human input.
Gmail-triggered n8n workflow: when a job enquiry email arrives, the Claude API extracts client name, job type, location, and urgency. A structured row is written to the jobs Google Sheet. An acknowledgement email is sent automatically. High-urgency jobs trigger an instant WhatsApp notification to the office manager. Weekly digest of open enquiries vs booked jobs sent every Monday morning.
An orthodontic practice with patients on 12–24 month treatment plans was relying on the front desk to manually send appointment reminders, follow up lapsed treatment plans, and request reviews after brace removal. With 200+ active patients and a small admin team, things were falling through — no-show rates were at 18% and review collection was inconsistent.
A multi-stage patient journey workflow triggered from their Google Sheets patient tracker: appointment reminders via WhatsApp at 48hrs and 24hrs, automatic treatment plan check-in messages at monthly intervals, re-engagement sequences for patients who missed appointments, and a timed review request 2 weeks after brace removal — routed to Google Reviews.
A mid-size recruitment agency placing into financial services received an AI due diligence questionnaire from a major institutional client. Their AI-assisted CV screening and automated candidate ranking tools — in daily use — fell squarely into the EU AI Act's high-risk category. They had no documentation, no risk register, and no policy framework. The client deadline was four weeks away.
Full AI system inventory and EU AI Act risk classification. Gap analysis against ICO automated decision-making guidance and Data Use and Access Act 2025. Risk-rated findings report. AI governance policy, candidate transparency notice, supplier due diligence questionnaire, and a prioritised remediation roadmap — all delivered within three weeks.
A technical recruitment agency was receiving 80–120 CVs per role, with consultants spending 3–4 hours per vacancy on initial screening before a single call was made. Roles with tight deadlines were being lost to faster competitors. The MD wanted consultants doing what they were hired to do — building relationships — not manually triaging applications.
An n8n workflow with Claude API integration: inbound CVs via email are parsed automatically, scored against a role brief, and ranked with a plain-English summary for each candidate — skills match, experience gaps, and a recommended next action. A shortlist of the top 5 candidates is emailed to the consultant within minutes of the application arriving. Human review retained at every decision point.
This build was delivered with full EU AI Act compliance documentation — human oversight at every decision point, candidate transparency notices, and a risk register. Compliant by design, not retrofitted.
A registered manager at a 28-bed private care home was spending her evenings on family enquiry responses, chasing rota gaps, and manually pulling together CQC documentation before inspections. She estimated 14–16 hours per week on admin that didn't require her clinical judgment. Staff retention was also suffering — bank staff weren't being contacted promptly when gaps appeared.
Three connected workflows: (1) Family enquiry triage — inbound emails and web form submissions are summarised and categorised, with standard enquiries acknowledged automatically and complex ones flagged for personal response. (2) Rota gap alerts — gaps in the Google Sheets rota trigger WhatsApp messages to the relevant bank staff pool, with availability confirmation logged automatically. (3) CQC document workflow — on-demand generation of a structured evidence pack from existing logs and records, ready for inspection.
All three workflows connect to tools already in use — Gmail, WhatsApp, and Google Sheets. No new software subscriptions. No system migration. The manager and her team changed nothing about how they work.
A free 30-minute call. Tell us what's costing your team time or what compliance obligation you've been putting off. We'll tell you honestly whether we can help and what it would cost.