We brought mynovarix in for the anomaly detection project after struggling with a vendor that kept changing scope. The contrast was significant. Deliverables were defined up front, the timeline held, and the system has been running without issues since deployment. The 30-day review period gave our team time to ask questions properly.
The speech recognition integration covered three languages — Bahasa Malaysia, English, and Mandarin — which was the main reason we chose mynovarix. That part worked well. The API integration took slightly longer than expected but the team communicated clearly throughout and the final result handles our call centre audio reliably.
We ran the strategy workshop series before committing to any implementation. It was exactly the right move. Siti led the sessions and each one ended with a concrete document we could actually use. Our leadership team came out of it with a shared understanding of where AI fits — and where it doesn't — in our business.
The published fixed pricing made budget approval straightforward. With most AI vendors, you go through several rounds of scoping before getting a number. mynovarix's approach saved weeks of internal process. The anomaly detection work itself was thorough — proper baseline modelling rather than just threshold rules.
We have used three different AI consultancies over the past four years. What sets mynovarix apart is the documentation. When the engagement finished we had everything we needed to maintain and extend the system ourselves. Most firms deliver something that works but can only be sustained with their continued involvement.
The workshop series was run for our leadership team of nine people. Siti managed the group well — different levels of AI familiarity in the room, and she adjusted the pace and depth accordingly. By the fourth session we had a prioritised list of AI initiatives ranked by feasibility and business impact. Very practical outcome.
A Selangor-based manufacturer was experiencing unplanned downtime due to equipment failures that sensor data in theory should have predicted. Their existing monitoring setup generated too many low-quality alerts and the operations team had learned to ignore them.
mynovarix deployed the Anomaly Detection Framework over eight weeks, building behaviour baselines for the specific equipment types involved. Alerting was reconfigured to surface only statistically significant deviations, with severity tiering to help the operations team prioritise.
Alert volume dropped by 71% while the proportion of alerts leading to a preventive action increased from roughly 12% to 58%. Three months post-deployment the client reported no unplanned downtime attributable to the monitored equipment categories.
A Kuala Lumpur-based financial services firm recorded thousands of customer service calls monthly but had no way to make the content searchable or usable for compliance review. Manual sampling covered only a fraction of recordings.
The Speech Recognition Integration was configured for the client's terminology mix — financial vocabulary in English, Malay, and Mandarin. A batch transcription API was developed to process the historical archive alongside real-time routing for new recordings.
Transcription accuracy for their specific terminology reached 91%. Compliance review time for sampled calls decreased by approximately 60% due to searchable transcripts. The archive of 14 months of historical recordings was processed in under two weeks.
An initial conversation takes 30 minutes and carries no obligation. Tell us about your situation and we will give you an honest view of how we can help.
Get in Touch