Farnagates

Work

Selected case studies.

Anonymised by default. Full references are available on request.

Case study 01

A UK water-services company

Five separate systems, decades of scanned paperwork and marketing spend that was hard to tie to revenue, brought together step by step so they now work as one.

Sector

Water services

Borehole drilling, ground-source heat pumps, ongoing maintenance

Size

Owner-managed

Small to mid-sized, multiple field teams

Engagement

Ongoing

Multi-phase, priorities agreed each cycle

Scope

Business-wide

Estimating · geology · operations · marketing · AI

The situation

A long-established services business with strong technical expertise but software that had grown organically over years. Quoting lived in a spreadsheet. Project paperwork lived in shared drives nobody fully trusted. Historic engineering reports, genuinely valuable data, were locked inside thousands of scanned PDFs. The website worked, but Google Ads and search performance were a black box: money was going out, leads were coming in, and the link between the two was guesswork.

The challenges

  • Quotes in spreadsheets, projects in Monday, scheduling in Teamup, accounts in Sage, HR in BreatheHR: five systems, none talking to each other.
  • Producing a quote was slow and inconsistent, because every estimator did it slightly differently.
  • Decades of valuable engineering reports were trapped in scanned paperwork. Re-using that knowledge meant rummaging through filing cabinets.
  • Google Ads spend could not be tied back to actual won jobs, so the bidding had no idea which campaigns were worth more.
  • Time-tracking ran on a fragile spreadsheet workflow disconnected from BreatheHR, with no audit trail.
  • When a job was sold, the handover to the delivery team relied on someone remembering to tell someone else.

What we did

Existing systems, connected across the whole business.

We started by spending time with the team to understand how the business really ran, then agreed priorities together. Work was delivered in short, visible steps, so something useful landed most weeks, and AI was introduced only where it clearly helped.

01

A custom estimator

Quotes built the same way every time, with the company's own pricing logic and a review step for senior staff.

02

Core systems connected

Project management, accounts, HR and the crews' calendar now share information. A won quote becomes a live project automatically.

03

AI that reads the archive

Decades of scanned engineering reports turned into a searchable knowledge base, with a person reviewing anything uncertain.

04

3D and machine-learning geology models

Models that combine public geological data with the firm's own drilling records to assess likely yield and water quality, and to design the borehole construction.

05

Servicing and operations in the field

Tablet forms for service visits, a planning view for the rigs and time-tracking linked to HR.

06

Marketing tied to revenue

A rebuilt website and connected search and advertising data, so the team can see which spend leads to real enquiries.

Systems and integrations

Connected to the tools the business already relied on.

Rather than replacing everything, the platform works alongside the existing systems and public data sources, so people keep using the tools they know.

Platforms and data connected

  • Monday.com
  • Teamup
  • BreatheHR
  • Sage
  • Microsoft 365
  • Google Ads
  • Search Console
  • British Geological Survey
  • Ordnance Survey

Where AI helps, with a person in the loop

  • Reads scanned reports and logs into structured, searchable data.
  • Drafts geology assessments for an engineer to check.
  • Drafts marketing content and summaries for human approval.

Results

What changed.

Five disconnected systems integrated, now working as one.

Decades of scanned engineering paperwork now searchable.

Marketing spend tied directly to won jobs.

  • Quoting time materially reduced and consistency restored, with every quote following the same logic.
  • A single, trusted view of every customer across sales, delivery and servicing.
  • Office staff freed from repetitive data entry, with the platform doing the heavy lifting.
  • A fortnightly delivery rhythm, with improvements landing continuously rather than as risky big releases.

What's next

The work continues in the same way it began: reviewing priorities with the team each cycle and building the next most useful piece. The longer aim is a single, continuous record of every customer, from first enquiry through quote, delivery and years of servicing.

Case study 02 · in progress

A UK finance-sector business

Currently in delivery. Three workstreams in parallel.

  • Process automation

    Automating repetitive, low-value processes across the back office.

  • Knowledge base

    Building a searchable knowledge base from decades of scanned and digital documents.

  • Compliance-ready documents

    Generating client documents automatically, in line with the firm's compliance procedures.

We work mostly under non-disclosure. If you'd like to talk to a previous client directly, we'll arrange it.

Worth a conversation?

An informal conversation about where things stand and what might help. No obligation, and no pressure to commit to anything.

Start a conversation