What Our Clients Say
Organizations across Hong Kong trust Tensorhive to deliver results. Read their stories.
Back to HomeClient Testimonials
James Tang
Legal Operations Director, Asian Law Partners
"The legal document review system cut our initial contract screening time by 60%. More importantly, the team understood exactly how the model worked—no black box. We've already planned a second engagement."
March 2026
Michelle Kong
Head of Planning, KMH Logistics
"Their forecasting work improved our demand projections significantly. The engagement was well-structured, the team was responsive, and the final dashboard makes it easy for our team to act on the predictions."
March 2026
Robert Morris
VP Risk & Compliance, Pan-Asia Finance
"The model audit gave us confidence our risk-detection system performs fairly across customer segments. The audit report was thorough and the recommendations helped us improve model performance."
February 2026
Susan Lee
Operations Manager, Hong Kong Imports Ltd
"The forecasting model reduced our inventory write-offs by 15%. Sarah explained everything clearly, and our team feels confident maintaining and updating the system on our own."
January 2026
David Fung
Data Officer, Hong Kong Insurance Co.
"The audit identified gaps we hadn't considered in our claims model. The recommendations were actionable and specific. Tensorhive took time to understand our regulatory environment."
January 2026
Victoria Wong
CTO, Asia Tech Ventures
"We engaged Tensorhive to validate an internal ML effort. Their audit was rigorous and constructive—not just a checklist. The insights helped us strengthen our models significantly."
December 2025
Case Studies
Legal Document Review Accelerates M&A Due Diligence
Asian Law Partners, Corporate Law Firm
Challenge
8 weeks
Manual review time for large due diligence engagement
Solution
AI Model
Legal document review system
Result
3.2 weeks
Effective screening timeline (60% reduction)
Engagement Overview
A mid-sized law firm managing a major acquisition needed to review thousands of vendor contracts and service agreements. The manual process was bottlenecking the timeline. We trained a custom NLP model on their historical contract data and configured the system to flag problematic clauses and extract key terms.
Process
Discovery phase identified the firm's specific review criteria and flagging rules. Model training took four weeks using their contract library. We integrated the system with their document management platform, trained the legal team, and provided support during the engagement.
Outcome
The model correctly identified 94% of clauses that required attention, reducing manual screening time significantly. Senior attorneys focused on substantive analysis rather than document triage. The firm successfully completed the due diligence on schedule and is now using the system for ongoing contract management.
Time Series Forecasting Improves Inventory Planning
KMH Logistics, Hong Kong-based Import/Export
Challenge
18%
Forecast error rate, causing stockouts and overages
Solution
ARIMA + ML
Hybrid forecasting approach with dashboard
Result
6.2%
Error rate, inventory optimization achieved
Engagement Overview
KMH needed better demand forecasts across their product portfolio. High forecast errors were leading to both stockouts that damaged customer relationships and excess inventory that tied up capital. We developed a forecasting system that combines statistical methods with machine learning.
Process
We analyzed five years of sales history, identified seasonality patterns, and tested multiple forecasting approaches. The final model accounts for seasonality, trend, and promotional effects. We built a dashboard allowing planners to generate multi-horizon forecasts and manage exceptions.
Outcome
Forecast accuracy improved from 82% to 93.8%. The planning team can now generate reliable one-month, three-month, and six-month projections. This has reduced inventory carrying costs by 12% and improved service levels to key customers. The system is updated automatically with new sales data.
AI Model Audit Identifies Fairness Issues Before Scale
Pan-Asia Finance, Regional Financial Services
Challenge
Unknown
Potential bias in credit-scoring model
Solution
Audit
Comprehensive fairness & performance review
Result
7 Issues
Identified and resolved pre-deployment
Engagement Overview
Pan-Asia Finance had deployed an internal credit risk model to a small user base and wanted independent validation before wider rollout. They were particularly concerned about fairness across geographic regions and age groups.
Process
We reviewed the model architecture, training data composition, and feature engineering decisions. We conducted stratified performance testing across demographics and identified performance drift in underrepresented segments. Analysis revealed that certain features had unintended proxy effects.
Outcome
The audit identified seven issues that could affect fairness and performance. Working with the client's data science team, we recommended resampling strategies, feature adjustments, and calibration improvements. The revised model now shows consistent performance across segments. The firm gained confidence to scale the model to broader use with ongoing monitoring.
Our Track Record
18+
Projects Completed
4.7/5
Average Client Rating
94%
Model Accuracy Rate
100%
Data Privacy Compliance
Ready to Become Our Next Success Story?
Let's discuss what's possible for your organization. No commitment—just a conversation about your challenges and how we might help.
Start the Conversation