Client Feedback
What Clients Say About Working With Wrenth
Honest accounts from organisations across Malaysia that have completed AI engagements with us.
Back to Home40+
Engagements completed
4.8
Average satisfaction rating
94%
On-time delivery rate
5
Industries served in MY
Direct Feedback
Client Testimonials
Chong Li Fen
Operations Manager · Petaling Jaya
We had years of financial reports locked in PDFs that nobody had time to process manually. The extraction system Wrenth built handles about 300 documents a week now. The output accuracy was higher than I expected given how inconsistent our document formats were.
Table Extraction · March 2026
Rajan Subramaniam
CTO · Digital Media Startup, KL
Content moderation was a real bottleneck for us. The taxonomy work at the start of the engagement was more thorough than I expected — they actually pushed back on some of our category definitions because they would have made the model harder to train. That kind of direct input was valuable.
Content Moderation · February 2026
Nurul Faizah
Head of Data · Financial Services, Shah Alam
The data strategy engagement gave our leadership team a shared language for talking about data. Before this, different departments had very different ideas about what we had and what was usable. The inventory work was the most useful part — we found data we didn't know was being collected.
Data Strategy · March 2026
Ahmad Hazwan
Research Analyst · Putrajaya
I was cautious at first because we'd had bad experiences with vendors who overpromised. Wrenth was quite specific about what the system would and wouldn't handle well, which made it easier to trust the outcome. The six-week timeline held and the delivery matched what we'd agreed on.
Table Extraction · January 2026
Siti Aishah Che Hassan
Product Lead · E-commerce Platform, KL
The human review queue they set up was something we hadn't thought to ask for but immediately understood why it mattered. Our team has final say on anything flagged with lower confidence, which keeps us in control. I'd recommend starting with a moderation engagement even if you think you need something bigger — it clarifies a lot.
Content Moderation · February 2026
Tan Kah Wei
Director of Strategy · Property Firm, KL
We weren't sure if we were ready for AI at all. The data strategy engagement answered that question concretely — some areas yes, some not yet, and a clear priority list for what to fix first. The RM 1,200 fee was modest relative to the clarity it provided to our planning process.
Data Strategy · March 2026
Detailed Accounts
Success Stories
Case Study · Finance Sector
Automating Table Extraction for a Regional Credit Provider
Challenge
A credit provider in Klang Valley was manually re-keying data from hundreds of monthly financial statements. The process took a full-time team member two days each month and was prone to errors that only surfaced weeks later.
Solution
Wrenth built a document ingestion pipeline that identifies table regions across varied PDF layouts, extracts cell data with positional awareness, and outputs structured rows to a connected database. Edge cases with merged cells are flagged for human review.
Results
Manual data entry for this task dropped from two days to approximately two hours monthly. Accuracy improved from an estimated 93% to over 98.5% on verified documents. The pipeline has processed over 1,200 statements since delivery.
"It handled our older scanned PDFs better than we expected. Those were the ones we were most worried about."
— Operations Lead, client organisation
Case Study · Platform Industry
Building a Content Moderation System for a Malaysian Classifieds Platform
Challenge
A classifieds platform was seeing a growing volume of listings that violated their terms — including prohibited items and misleading product descriptions. Their moderation team was consistently backlogged, and review turnaround was affecting seller experience.
Solution
Wrenth designed a moderation taxonomy from the platform's existing policy documents, trained a classification model on labelled listing data, and integrated a review queue that routes low-confidence flagging to human reviewers before any action is taken.
Results
Approximately 78% of policy violations are now auto-flagged before reaching the review queue. Human reviewer time reduced by about 60% within the first month. False positive rate held below 4% across the first quarter of operation.
"The taxonomy workshop at the start was a bit uncomfortable because it forced us to be precise about things we'd left vague. But it made everything after that much smoother."
— Product Lead, client organisation
Case Study · Professional Services
Data Strategy for a Mid-Sized Consulting Firm Preparing for AI Adoption
Challenge
A consulting firm wanted to integrate AI into its research and report generation workflows but had no clear picture of what data they actually held, how it was stored, or whether it was in a usable state for any AI application.
Solution
Wrenth conducted a four-week data inventory exercise across their systems, scored data assets by quality and accessibility, and produced a strategy document with a 12-month roadmap prioritising the three highest-value improvements before any AI development began.
Results
The strategy document was presented to the firm's board and approved within two weeks. Three of the four recommended data improvements have since been implemented. The firm began a table extraction engagement with Wrenth in the following quarter.
"We found things during the inventory that nobody realised we had. It changed how we thought about what was feasible."
— Strategy Director, client organisation
Professional Standing
Certifications & Affiliations
MSC Status, Malaysia
Active · Ministry of Digital
MDEC Digital Services Recognition
March 2026
Cradle Fund CIP500 Participant
2024 cohort
Reach Out
Contact Wrenth
Add Your Organisation to the List
A short conversation is enough to understand whether there's a good fit. No obligation to proceed.
Get in Touch