What Technical Teams Say
Feedback from engineering teams and technical leadership who have worked with Quorux on AI implementation projects.
Back to HomeClient Feedback
Direct feedback from technical stakeholders on delivered projects and working relationships.
Wei Chen
ML Engineering Lead
Kuala Lumpur, Malaysia
The architecture work Quorux delivered for our computer vision pipeline addressed specific bottlenecks we had been struggling with for months. Their ablation studies clearly showed which components contributed to performance gains, and the documentation allowed our team to understand and modify the implementation independently.
January 28, 2026
Siti Abdullah
CTO
Penang, Malaysia
The integration consulting helped us avoid several expensive mistakes in our AI platform planning. Their infrastructure sizing recommendations proved accurate once we moved to production, and the vendor evaluation matrix gave us clear criteria for making informed decisions rather than relying on sales pitches.
February 5, 2026
Raj Kumar
Data Science Manager
Johor Bahru, Malaysia
Custom OCR work for our historical archive digitization project exceeded expectations. The system handles degraded Tamil and Jawi script documents that general-purpose OCR couldn't process. Accuracy improvements were measurable and the deployment package included everything we needed to integrate with our existing workflow.
January 19, 2026
Lee Ting
Senior Software Engineer
Kuala Lumpur, Malaysia
Working with Quorux felt like collaborating with experienced engineers rather than dealing with a vendor. They were transparent about what would work well versus what required significant effort, and they structured the engagement with checkpoints that allowed us to evaluate progress before full commitment.
February 11, 2026
Ahmad Hassan
Technical Director
Shah Alam, Malaysia
The technical depth was impressive. Rather than applying cookie-cutter solutions, they researched our specific problem domain and experimented with multiple architectural approaches. The final delivered model met our latency requirements while maintaining accuracy targets, which several other consultants had claimed wasn't possible.
January 23, 2026
Maya Ng
Product Manager
Petaling Jaya, Malaysia
The knowledge transfer was valuable. They didn't just hand over code — they explained the reasoning behind design decisions, pointed us to relevant research papers, and answered detailed questions from our engineering team. This built internal capability rather than creating vendor dependency.
February 2, 2026
Success Stories
Detailed case studies showing how Quorux solutions addressed specific technical challenges.
Logistics Route Optimization Model
Challenge
A logistics company needed to optimize delivery routes considering traffic patterns, vehicle capacity constraints, and time windows. Existing heuristic approaches weren't adapting well to changing conditions in Kuala Lumpur's road network.
Solution
Designed a graph neural network architecture that models the road network as a dynamic graph with time-varying edge weights. Incorporated attention mechanisms to weight route segments based on current traffic data and historical patterns.
Results
Achieved 18 percent reduction in average delivery time and 23 percent improvement in on-time delivery rates. The model adapts to real-time traffic conditions while respecting all operational constraints. Delivered in three weeks.
Financial Services AI Platform Planning
Challenge
A bank wanted to integrate AI capabilities for fraud detection and risk assessment but faced uncertainty about infrastructure requirements, vendor selection, and regulatory compliance implications. Existing IT architecture needed careful assessment.
Solution
Conducted comprehensive infrastructure assessment including data pipeline requirements, latency constraints, security framework compatibility. Evaluated multiple vendors against specific criteria. Developed phased implementation roadmap with risk mitigation strategies.
Results
Client avoided significant infrastructure overspending by right-sizing compute resources. Vendor selection criteria led to choice that met technical requirements at 40 percent lower cost than initially budgeted. Implementation proceeded on schedule without major technical surprises.
Government Archive Digitization
Challenge
A government agency needed to digitize historical documents including multilingual content in Malay, Tamil, and Jawi script. Many documents showed degradation from age and storage conditions. General OCR systems achieved below 60 percent accuracy.
Solution
Trained specialized OCR models on annotated samples of their specific document types. Expanded character sets to include all script variants. Developed preprocessing pipeline to handle degradation and layout analysis tuned for their document formats.
Results
Achieved 91 percent character-level accuracy on test set, up from 58 percent with general OCR. Processing speed of 45 pages per minute met throughput requirements. System deployed and operational within two and a half weeks from project start.
Trust Indicators
Measurable performance metrics and professional credentials.
Client Satisfaction
Based on post-project surveys
Projects Delivered
Across multiple sectors
Organizations Served
Government and private sector
Years Operating
Since January 2021
Professional Credentials
ISO 27001 Compliance
Information security management standards
ACM Professional Member
Association for Computing Machinery
IEEE Computer Society Member
Focus on AI research and applications
MDEC Recognized Provider
Malaysia Digital Economy Corporation
Join Our Growing Client Base
Work with engineers who deliver measurable results through careful technical work and honest assessment.
Start Your Project