What Participants Say About the Programmes
Feedback from analysts, developers, and team leads who've attended Voltcortex cohorts and labs in Kuala Lumpur.
← Back to HomeFrom People Who've Attended
"I'd been working alongside ML engineers for two years without understanding why they kept mentioning VRAM limitations. The Foundations cohort gave me the mental model I was missing. The evening format meant I didn't have to take any leave."
"The lab session was the useful part. Running our own jobs and watching the memory counter climb in real time is something you can't replicate from a tutorial. The facilitator was patient when we asked questions that were probably obvious to them."
"We brought five people from our data engineering team through the Team Enablement. The intake process was genuinely useful — they asked about our actual infrastructure decisions, which made the exercises feel relevant rather than generic."
"I was sceptical about whether a four-evening programme could actually change how I think about compute. It did. The analogies they use for parallelism are well-designed — I've already used them twice when explaining GPU costs to stakeholders."
"The tuning reference sheet has come up several times at work since the lab. It's not something I would have thought to produce myself — it's a practical shorthand for things we kept having to look up. Minor note: the lab days are long; bring lunch."
"Our organisation was about to start evaluating cloud GPU options and nobody on the team had a consistent picture of what mattered. After the Team Enablement, our internal discussions were noticeably more specific. The handbook is still on our shared drive."
How Organisations Have Used the Programmes
Starting Point
A Kuala Lumpur-based payments company was planning to move model inference from a CPU-based service to GPU-hosted endpoints. The engineering and product teams had different assumptions about what this would mean for latency and cost. Procurement discussions kept stalling because participants were talking past each other.
What We Did
Seven people — three engineers, two product leads, and two infrastructure stakeholders — went through the Team Enablement Programme over six weeks. The intake session mapped their specific decision: single-GPU endpoints vs. multi-GPU batch inference. Sessions were anchored to that context throughout.
After the Programme
The team completed their evaluation process within four weeks of the programme ending. In their words: discussions about memory requirements and throughput targets now happen in a shared vocabulary rather than a negotiation over assumptions. The handbook remains in active use.
Starting Point
A regional e-commerce group had expanded into ML-assisted demand forecasting. Their data team's developers understood the models but not the infrastructure; their platform engineers understood the infrastructure but not the models. Neither group could have a complete conversation about GPU provisioning.
What We Did
Three product analysts attended the Foundations cohort to build conceptual grounding. Separately, four developers attended a Compute Lab session to gain hands-on experience with memory-bound jobs. Both groups received the glossary, giving them a shared reference despite attending different programmes.
After the Programme
Cross-functional reviews of GPU usage — previously avoided because of knowledge gaps — began appearing on sprint agendas. The team reported that job sizing decisions, which had previously been delegated entirely to a single engineer, were now being made collaboratively with a wider group.
By the Numbers
Reach Out
Bangsar South, 59200 Kuala Lumpur
Saturday: 10:00 am – 2:00 pm
Ready to Join the Next Cohort?
Whether you're registering for yourself or arranging a programme for a team, get in touch and we'll walk you through the options and upcoming schedule.