Testimonials · Client Experiences
What Clients Have Found
Accounts from organisations that have worked with Quantellis across feasibility, prototyping, and knowledge transfer engagements.
← Return to HomepageEngagements completed
Average client rating
Return engagement rate
Sectors served
§ 1 — Client Reviews
In Their Words
Roshani Hassan
Head of Operations · Logistics firm, Shah Alam
We engaged Quantellis for a feasibility study on route optimisation. What I valued most was that they told us, clearly, which parts of our problem were well-suited to AI and which were not. That honesty saved us from an expensive project that would not have delivered what we hoped.
February 2026 · Feasibility Study
Kenneth Wong
CTO · Financial services company, KL
The prototyping work Quantellis did for our credit risk team was genuinely useful. The benchmark report in particular — showing exactly where the model performed and where it struggled — gave us what we needed to make an informed decision about next steps. The documentation quality was high.
January 2026 · Prototyping
Nurul Zainab
Analytics Manager · FMCG company, PJ
Our team attended the knowledge transfer programme after a prototyping engagement. The training used our actual data, which made an enormous difference. We came out of it with a clearer understanding of what we had built and how to work with it. Karthik was a patient and thorough facilitator.
March 2026 · Knowledge Transfer
Arjun Sharma
VP Technology · Healthcare group, Subang Jaya
I had worked with a few AI consultancies before Quantellis and the difference is in the rigour. They do not start with a solution and work backwards — they actually investigate the problem first. The feasibility report we received was the most substantive document of its kind I have read from any consultant.
February 2026 · Feasibility Study
Lim Mei Ling
Digital Director · Retail group, Bangsar South
The prototyping engagement for our demand forecasting problem was handled with a level of care I had not expected from a small consultancy. What I appreciated was the honesty about the prototype's limitations — they did not oversell what it could do. That gave us confidence in the parts that did work well.
January 2026 · Prototyping
Farouk Ibrahim
Head of R&D · Manufacturing, Shah Alam
We used Quantellis for both a feasibility study and, after the study recommended it, a prototype. The continuity between the two phases was valuable — the same consultant, the same understanding of our context. We did not have to re-explain anything.
March 2026 · Feasibility + Prototyping
§ 2 — Case Studies
Selected Engagement Summaries
Evaluating AI for Delivery Route Optimisation
Challenge
A mid-sized logistics company in the Klang Valley wanted to evaluate whether machine learning could reduce fuel costs and delivery times by improving route planning. They had route history data but no AI experience.
Approach
Quantellis conducted a four-week feasibility study reviewing relevant optimisation approaches, assessing the client's route history data quality, and modelling what improvement was realistically achievable.
Outcome
The study concluded that AI-based optimisation was feasible for long-haul routes but not short urban runs with high variability. The client used this to target a prototype for the specific context where it made sense.
"The study gave us a credible, specific answer — not a vague 'yes, AI can help'. That is what we needed."
— Head of Operations, Shah Alam · February 2026
Document Classification Prototype for Compliance
Challenge
A Kuala Lumpur financial services firm was manually classifying large volumes of compliance documents. The process was time-consuming and prone to inconsistency across different staff members.
Approach
A six-week prototyping engagement developed an NLP-based document classification model trained on three years of labelled compliance records. Data preparation took approximately half the total project time.
Outcome
The prototype achieved 87% classification accuracy on the test set, with documented failure patterns for the remaining 13%. The client proceeded to a production build with a third-party developer using the prototype as a reference.
"Knowing exactly where the model struggled was as valuable as knowing where it worked. The benchmark report made our follow-on decisions straightforward."
— CTO, KL Financial Services · January 2026
Building Internal AI Capability for Quality Control
Challenge
A Shah Alam manufacturer had a small data team that understood statistics but lacked applied ML experience. They wanted to develop their own AI models for quality control rather than depending on external vendors.
Approach
A two-module knowledge transfer programme was designed around the team's specific gap: supervised learning for tabular sensor data. Sessions used the company's own quality records throughout, rather than standard teaching datasets.
Outcome
The team developed and deployed their first internal anomaly detection model within eight weeks of the programme completing. Follow-up mentoring sessions over four weeks supported the transition to independent project ownership.
"The difference from other training is that we were working with our real data from day one. Everything we learned had immediate context."
— Head of R&D, Shah Alam Manufacturing · March 2026
§ 3 — Contact
Reach Us Directly
Telephone
+60 3-4167 8253Address
8 Jalan University, 46200 PJ
Hours
Mon–Fri, 9:00–18:00
Next Step
Would you like to discuss your own situation?
An initial conversation is informal and without obligation. We are happy to share our honest view of whether our services would be useful to you.
Get in Touch