Case studies
Results we can put a number on.
Five engagements across our three practices — each one built around a number the business needed to move.
Industrial Operations
From 2.5 days to 4 hours: automating quoting for an industrial distributor.
The challenge
Sales engineers spent most of each day reading requests for quotes from emails and PDFs, checking stock and pricing in the ERP, and rebuilding quotes by hand. Quotes took days. In a market where the first credible quote usually wins, that delay was costing the business orders.
What we did
- 1Mapped the quoting process end to end and agreed the baseline: turnaround time, win rate and engineer hours per quote.
- 2Built an AI agent that reads incoming requests, matches parts to the catalogue and pulls live pricing and stock from the ERP.
- 3Generated complete draft quotes for engineers to review and send — people kept the final decision on price and terms.
- 4Rolled out across all 6 branches with weekly reporting against the baseline.
The result
4 hrs
Quote turnaround
from 2.5 days
26%
Quote win rate
from 18%
−60%
Engineer hours on quoting
per week
$2.1M
Additional revenue won
in the first two quarters
“Our engineers used to spend their mornings retyping RFQs. Now they review a finished quote and get back to customers the same day — and we’re winning work we used to lose on speed.”
Healthcare
Becoming the practice AI assistants recommend — and filling the schedule.
The challenge
Patients were increasingly asking ChatGPT and Google's AI answers which provider to see — and the group rarely appeared. Paid search costs were climbing, and online enquiries waited hours for a reply, so many booked elsewhere.
What we did
- 1Audited visibility across ChatGPT, Perplexity, Gemini and Google AI Overviews for 40 priority treatments and locations.
- 2Restructured site content, listings and review signals so AI systems could clearly identify, trust and cite the group.
- 3Rebuilt paid campaigns to optimise for booked appointments rather than clicks.
- 4Added an AI assistant that answers enquiries and books appointments around the clock, handing complex cases to staff.
The result
27 of 40
Priority AI queries citing the group
from 3 of 40
+34%
New patient bookings
vs. same period last year
−28%
Cost per booked patient
from $142
<1 min
Enquiry response time
from 4.5 hours
“We went from invisible in AI answers to being the practice it recommends in our markets. New patient volume is up, and we’re paying less for every booking.”
Agriculture · Carbon markets
Building intelligence into agricultural carbon operations.
Carbon Cloud's technology roadmap goes beyond traditional project management to bring in intelligent automation, predictive analytics and AI-assisted operational decision-making. The platform architecture is built so these capabilities can be introduced as the business grows.
The AI architecture
Designed around five capabilities.
- 01Intelligent operations agentsMonitor project activity, flag incomplete workflows, surface operational exceptions and recommend next actions across the carbon project lifecycle.
- 02Predictive logistics intelligenceMachine learning that analyses material movements, transport activity, delivery schedules and field operations to spot likely delays and improve coordination.
- 03Automated environmental evidence processingExtracts, classifies and validates information from laboratory reports, field records, transport documentation and other project evidence.
- 04Geospatial & agronomic intelligenceAnalyses soil characteristics, geographic boundaries, field-level measurements and agricultural datasets to support project planning and environmental monitoring.
- 05AI-powered operational insightsNatural-language analytics and reporting, so business users can explore performance and operational risk without assembling data from multiple systems.
The business opportunity
Administrative workload
Less manual chasing, checking and re-keying.
Operational responsiveness
Issues and delays surfaced before they escalate.
Evidence quality
Validated, consistent project documentation.
Portfolio scale
More projects managed without proportional overhead.
“[Carbon Cloud quote on why this roadmap matters to the business.]”
Home Services · HVAC
Every call answered: an AI dispatcher for peak season.
The challenge
In summer, call volume more than tripled. Nearly a third of calls went unanswered during evening peaks and after hours — and in Phoenix in July, a homeowner without cooling calls the next contractor on the list. Customer service staff spent their days juggling the schedule instead of serving customers.
What we did
- 1Analysed 12 months of call recordings and job data and agreed the baseline: answer rate, call-to-booking rate and revenue per call.
- 2Deployed an AI voice agent that answers every call, triages it (emergency, repair, maintenance) and books directly into ServiceTitan by technician skill, zone and availability.
- 3Built emergency escalation to the on-call technician, with the customer’s history and equipment attached to the job.
- 4Automated follow-up on unaccepted estimates and maintenance-plan renewals by text, with staff handling anything the agent can’t.
The result
99%
Calls answered
from 71%
62%
Calls converted to booked jobs
from 48%
$940K
Additional revenue
over the first summer season
−45%
Staff hours on scheduling
per week
“In July we used to lose calls every single evening. Now every call gets answered, emergencies reach a technician in minutes, and the board is full before we open.”
Restaurants & Hospitality
More private events booked — and less food thrown away.
The challenge
Private dining and catering enquiries arrived by phone, email, Instagram and web form, and managers answered them between services — often a day later, by which time the customer had booked elsewhere. In the kitchen, prep was planned on instinct, so slow nights meant waste and busy nights meant running out.
What we did
- 1Agreed the baseline: enquiry response time, enquiry-to-booking rate, event revenue, and food waste as a share of food cost.
- 2Deployed an events assistant that replies on every channel within minutes with availability, menus and pricing, and drafts proposals for a manager to approve.
- 3Built a prep forecast from POS history, reservations, weather and local events that sets each location’s daily par levels by dish.
- 4Set up a weekly owners’ report on event pipeline, covers, food cost and waste by location.
The result
6 min
Event enquiry response
from 19 hours
37%
Enquiries booked
from 22%
+$410K
Event & catering revenue
annualised
−31%
Food waste
5.4% of food cost, from 7.8%
“Our managers used to answer event emails at midnight. Now every enquiry gets a proposal in minutes, and our kitchens prep for the night we’re actually going to have.”
What number do you need to move?
Book a free strategy session. We'll show you where AI would pay first in your business.