FONCIERE CINQ TERRES
AI Software Solutions — Verified Outcomes, Transparent Methods
Narrative Index
Case 01: Freight Document Automation
A Glasgow-based logistics company processed over 1,200 shipping manifests per week by hand. Errors averaged 9.4% monthly. Assurance AI Soft deployed a document-parsing pipeline using custom OCR and classification models trained on the client's own data formats.
Within 38 days, error rates dropped to 1.2%. The operations team reallocated 14 staff hours per day from data entry to exception handling and client liaison. The system paid for itself in the second month.
Verified outcome — Q3 2025 engagement. Client reference available on request.
Impact Snapshot
Case 02: Clinical Note Summarisation
The Challenge
An NHS-adjacent private clinic in Edinburgh needed to summarise patient consultation notes for referral letters. Clinicians spent an average of 22 minutes per referral composing summaries from unstructured notes.
What We Built
A fine-tuned NLP summarisation model integrated directly into the clinic's existing EHR interface. The model was trained on anonymised historical notes with clinician review loops built into the pipeline.
The Outcome
Referral letter composition time dropped to 7 minutes on average. Clinician satisfaction scores for the tool reached 8.6 out of 10 within the first quarter. No patient data ever left the clinic's infrastructure — all processing ran on-premise.
Engagement completed February 2025. Anonymised metrics shared with permission.
Case 03: Demand Forecasting for Independent Retail
A chain of 11 independent homeware shops across central Scotland struggled with overstocking seasonal items. Assurance AI Soft built a demand forecasting model using three years of POS data, local event calendars, and weather pattern feeds.
"We reduced dead stock by 41% in the first season. The model even flagged a trend we hadn't spotted — outdoor lighting demand spikes two weeks before the Edinburgh Festival." — D. Rennie, Buying Manager
Stock holding costs fell by £18,200 across the group in six months. The model continues to learn from each season's data.
Capability Map — What We Engineer
| Capability | Typical Application | Delivery Model | Proof Available |
|---|---|---|---|
| Document Intelligence | OCR, parsing, classification of business documents | On-premise or private cloud | Case 01 |
| NLP & Summarisation | Clinical notes, legal briefs, support tickets | On-premise preferred | Case 02 |
| Predictive Analytics | Demand forecasting, churn prediction, resource planning | Cloud or hybrid | Case 03 |
| Computer Vision | Quality inspection, asset monitoring, safety compliance | Edge or on-premise | Case 05 |
| Conversational AI | Internal knowledge bots, customer-facing assistants | Cloud with data residency | On request |
| Data Pipeline Design | ETL, feature engineering, data quality automation | Hybrid | All cases |
Decision Board — Is AI Right for Your Problem?
Good Fit
You have a repeatable process with structured or semi-structured data. Volume is high enough that manual handling creates bottlenecks. You can define what "better" looks like — fewer errors, faster turnaround, lower cost per unit.
You have at least six months of historical data, even if it's messy. You have internal stakeholders willing to validate model outputs during a pilot phase.
Not Yet Ready
Your process changes fundamentally every quarter. You lack any historical data to train on. Success criteria are undefined or purely subjective. There is no internal champion to own the project alongside our team.
This isn't a rejection — it's a timing question. We often help organisations reach readiness before starting a build.
Case 04 & 05 — Condensed Narratives
A mid-tier law firm in Aberdeen needed to review lease agreements faster during a commercial property portfolio acquisition. We built a clause extraction and risk-flagging tool that reduced first-pass review time from 4 hours to 45 minutes per document. The tool flagged 12 non-standard indemnity clauses that the manual review had missed in a previous transaction. Engagement ran for 9 weeks. The firm now uses the tool as standard practice for due diligence work.
A renewable energy cooperative managing 23 solar and wind installations across the Highlands needed better load prediction to reduce grid penalties. Assurance AI Soft deployed a time-series forecasting model fed by weather APIs, historical generation data, and National Grid demand signals. Grid penalty charges dropped by £31,400 over eight months. The cooperative's grid operator described the model as "the most useful tool we've adopted in five years." Computer vision was later added to monitor panel degradation from drone imagery.
Every engagement begins with a scoping conversation — not a sales pitch. We define the problem, the data landscape, the success metric, and the deployment constraint before writing a single line of code. We build pilot-first: a working model on real data within weeks, not months. We train your team to own the system. We do not create dependency. If the problem doesn't suit AI, we say so. Roughly one in five initial enquiries results in a recommendation to pursue a non-AI solution instead.
Start a Conversation
Find Us
242 Sporer-Gibson Lea, Old Green, Scotland, XT65 9WR, United Kingdom
Privacy Policy
Assurance AI Soft ("we", "us") collects only the personal data you voluntarily submit through our enquiry form: name, email, company, and message content. We use this data solely to respond to your enquiry and assess project fit. We do not sell, share, or transfer your data to third parties except where required by law. Data is stored securely on servers within the United Kingdom. You may request deletion of your data at any time by emailing [email protected]. We use a single localStorage flag to remember your cookie consent preference; no tracking cookies are deployed. This policy was last reviewed in January 2026.
Terms of Service
By using this website you agree to the following terms. The content on assuranceaisoft.sbs is provided for informational purposes. Case study metrics reflect specific client engagements and are not guarantees of future results. All intellectual property on this site belongs to Assurance AI Soft unless otherwise stated. We reserve the right to modify these terms at any time. Continued use of the site constitutes acceptance of any changes. Enquiries submitted through the contact form do not constitute a contractual agreement. Formal engagements are governed by separate project agreements. These terms are governed by the laws of Scotland.
Disclaimer
The case studies, metrics, and outcomes described on this website are based on specific client engagements conducted under particular conditions. Results vary depending on data quality, organisational readiness, and project scope. Assurance AI Soft does not guarantee identical outcomes for new engagements. The information on this site does not constitute professional advice. For decisions involving significant business investment, we recommend a formal scoping consultation. External images used on this site are sourced from placeholder services for illustrative purposes and do not depict actual client premises or personnel unless explicitly stated.