- Malaysian AI platform transforms familiar files into intelligent insights with data sovereignty
- LLM agnostic platform lets govts, enterprises keep apps, data, even GPUs within jurisdiction
A Malaysian startup is wagering that the bottleneck in enterprise AI isn’t models, it’s ingestion. Nervesis Sdn Bhd’s new platform, Zygy, is built to turn messy spreadsheets and contracts into AI-readable data, paired with hard controls to keep information inside national borders. It made its public bow during a livestream chat between founder-CEO Azhar K Mustapha and DNA’s Karamjit Singh at MDX Summit 2025 in mid Sept and signals Malaysia’s intent to compete with global tech companies on both performance and data sovereignty.
Early customer engagement around Zygy ranges from compliance checks ensuring vendors meet contract requirements for a near-closed Dubai deal, to powering a command-and-control centre for a defense contractor in Malaysia, to a local bank pilot that helps research analysts build financial indicators faster.
For smaller firms drowning in Excel and scanned docs, Zygy starts at only US$9.99 (RM42.20) a month with light analytics and Google Drive/OneDrive hooks. “There’s just too much to do. Without AI, they’d fall behind,” opines Azhar on the workload at smaller companies which always keep headcount tight to manage expenses and as a result are not likely taking advantage of the raw data they have.
Why ingestion matters and why being LLM-agnostic is a differentiator
“Ingestion is the first step. If you don’t get it right, the AI can’t respond intelligently,” Azhar argues. He explains that Zygy converts raw files into structured, AI-readable formats to improve chatbot responses, analytics, and decision support—born out of Nervesis’ own experience with early projects where existing tools struggled with complex spreadsheets.
Zygy is pitched as LLM-agnostic, meaning it can work with open-source models or region-specific providers like Mistral in France or DeepSeek in China, and orchestrated through AI agents, letting governments and enterprises keep apps, data, and even GPUs within jurisdiction. “In Dubai, they want their data to stay in Dubai. In Malaysia, it stays here. That’s what we enable.” Transparency around model training and fine-tuning is a stated selling point as well.
Founder’s edge as a MIT trained software engineer
Azhar is no stranger to the world of AI. Armed with a degree in computer science and electrical engineering from the Massachusetts Institute of Technology, he cut his teeth after graduation working at SpeechWorks, the Boston company whose voice technology later evolved into Apple’s Siri. He returned to Malaysia in the early 2000s, first as an academic at Multimedia University before venturing into entrepreneurship. By 2011, he was already building AI solutions for Telekom Malaysia.
That long track record in AI underpins Nervesis and now Zygy – proof that this isn’t just another opportunistic pivot into AI, but a continuation of decades of applied expertise.
The build and the bet
Nervesis runs lean with 10 people, seven of whom are technical and has between RM4 million to RM5 million in sales annually. Zygy’s go-to-market is cloud subscriptions and social outreach, with ambition set not against local rivals but global incumbents. “We’re not competing with local companies. Our goal is to be among the top 20 globally in this category,” says Azhar.
If Zygy truly nails ingestion on real-world spreadsheets and contracts — while proving sovereignty and transparency are more than slideware — Malaysia will have a credible AI export, not just another domestic demo. The proof will be in live deployments, not pitch-deck benchmarks. With promising early signs, Azhar is confidence that Zygy will deliver.
Learn more about Zygy at www.zygy.com
