Client Perspectives
What Clients Say
After the Work Is Done
The feedback below comes from organisations in Singapore that have completed engagements with Pulsemind. We've chosen to share a range of perspectives — including where things were harder than expected.
← Back to HomeReviews
Client Feedback
Tan Wei Jie
Head of Compliance, Financial Services
"We engaged Pulsemind for a governance framework ahead of an MAS technology risk review. The manual they delivered was genuinely usable — our legal team reviewed it without needing to simplify it first. The stakeholder workshops were well-structured and surfaced issues we hadn't considered."
January 2025 · Ethics & Governance
Priya Krishnan
Data Science Lead, Healthcare Admin
"The MLOps platform has genuinely changed how our team works. We'd been stuck in notebook-based experimentation for two years. The onboarding sessions were patient and thorough — Rajan took the time to understand how we actually work before suggesting a setup, which I appreciated."
February 2025 · ML Ops Platform
Marcus Chew
IT Director, Logistics Operations
"The anomaly detection system took a bit longer to stabilise than we expected — the first month of tuning surfaced some data quality issues we had to fix on our side first. Once that was sorted, the system performed well. The team was transparent about what was causing delays, which made the process manageable."
December 2024 · Anomaly Detection
Siti Lailah
AI Strategy Manager, Retail Group
"Jamie and the team made the governance workshops accessible for people who weren't technical. Our finance director, who was initially skeptical of the process, ended up being the most engaged participant. The training materials have since been used in our board-level AI briefings."
January 2025 · Ethics & Governance
David Ho
CTO, Fintech Startup
"We needed a transaction anomaly detection system that could differentiate between fraud signals and normal high-volume periods. Pulsemind understood the distinction immediately. The baseline modelling work was more sophisticated than I expected at this price point."
February 2025 · Anomaly Detection
Ng Li Ying
Senior Data Analyst, Insurance
"The runbooks are genuinely detailed — not just for show. Six months after the MLOps engagement ended, a new team member was able to use them to set up a new model pipeline without needing to contact Pulsemind. That's the test I would apply to any documentation."
November 2024 · ML Ops Platform
Case Studies
Three Engagements in Detail
Case Study 01 · Governance Framework
Challenge
A mid-sized financial services firm was expanding its use of AI in credit assessment and customer communication. Their compliance team had no formal framework for evaluating whether new AI tools met ethical and regulatory standards before deployment.
Solution
Pulsemind conducted a four-week engagement covering a review of current AI use cases, stakeholder workshops with compliance, product, and legal teams, and the development of a governance manual aligned to MAS Notice 688 guidance.
Outcome
The firm completed its MAS technology risk assessment with governance documentation in place. The review process developed during the engagement was subsequently applied to three new AI procurement decisions within six months.
"The process forced us to have conversations we had been putting off. The framework gave those conversations a structure they previously lacked."
— Compliance Director, Financial Services
Case Study 02 · Anomaly Detection
Challenge
A logistics operator was receiving alerts from a basic threshold-based monitoring system that generated a high volume of false positives, desensitising the operations team to genuine issues. A missed equipment fault had resulted in a costly service disruption.
Solution
Pulsemind designed an adaptive anomaly detection system that integrated with four existing telemetry data sources and established dynamic baselines per asset type and operational period. The two-month tuning period addressed seasonal and shift-based patterns.
Outcome
False positive volume reduced substantially in the first six weeks of tuning. The operations team reported a meaningful improvement in their ability to distinguish signal from noise. No service disruptions attributable to missed detection in the following four months.
"We'd had systems that flagged everything and systems that flagged nothing. This one actually flagged the right things."
— Operations Technology Manager, Logistics
Case Study 03 · ML Operations
Challenge
A healthcare administration provider had a data science team of four people building models in notebooks with no version control, no experiment tracking, and no reliable way to reproduce previous results. Model maintenance was consuming a disproportionate share of the team's capacity.
Solution
Over ten weeks, Pulsemind architected and deployed an ML operations environment incorporating version control, experiment tracking, and a continuous training pipeline. Four team onboarding sessions ensured the team could operate and adjust the setup independently.
Outcome
Model maintenance time reduced considerably within the first month. The team reported being able to run structured experiments for the first time. A new model was deployed to production within eight weeks of the engagement closing, using the established pipeline.
"The operational runbooks meant that when one of our senior data scientists left three months later, we didn't lose institutional knowledge along with them."
— Head of Data Science, Healthcare Administration
In Numbers
What the Record Shows
6+
Years Active
Established in Singapore in 2019
80+
Completed Engagements
Across Singapore and the ASEAN region
4.8
Average Rating
Post-project client satisfaction surveys
3
Core Services
Focused depth over broad coverage
Your Turn
Ready to discuss your situation?
We approach every initial conversation without obligation or pressure. If we can help, we'll say so. If we can't, we'll say that too.
Get in Touch