The appointment and what TracPlus actually runs
TracPlus named Audrey Cheng as chief technology officer on August 13, putting her in charge of engineering for the company's Mission Platform, the operational data layer underneath aerial firefighting operations in more than 40 countries. The platform serves as the system of record for major agencies including CAL FIRE in the United States, Australia's national aerial firefighting program, and New Zealand's Fire and Emergency service, managing approximately 2,500 wildfire suppression aircraft and holding more than one billion flight records accumulated over the platform's operating history.
That scale is not abstract: TracPlus processes roughly one million new position reports every day, each one a real-time signal that has to be integrated with operational, environmental, and geospatial data sources fast enough to support decisions being made by agencies actively fighting fires. This is infrastructure where latency and reliability failures have physical safety consequences, which puts Cheng's engineering mandate in a different risk category than most SaaS platforms.
A product and payments background applied to a safety-critical domain
Cheng's prior roles give her an unusual mix of relevant experience. As VP of Product at Pushpay, a payments platform that scaled through a genuine hypergrowth phase, she worked on the exact problem of keeping product velocity and system reliability aligned while user volume compounds quickly, a discipline that transfers directly to TracPlus's own scaling curve. Her time as Chief Product Officer at SnapComms, later acquired by critical events company Everbridge, and as CTO at Imagr adds direct experience in both crisis communications infrastructure and computer-vision-driven operational systems.
That combination, payments-grade reliability discipline plus critical-events domain experience, is a more specific fit for TracPlus's problem than a generic infrastructure hire would be. Scaling a payments platform and scaling a wildfire operations platform share the same underlying requirement: the system has to be right and fast under load, because the cost of a wrong or slow answer during an actual emergency lands on a decision made with incomplete information in a fire zone, not on a support ticket. That is a much higher bar for reliability engineering than most consumer software ever has to clear.
The stated problem is decision support, not data volume
Cheng was direct about where she sees the real gap: right now, so much of the best decision-making in agencies sits in experience, and what her team is building is transparency into what factors into those decisions. That framing matters because it draws a clear line between two very different engineering priorities, simply moving more data faster through the pipeline, versus making the reasoning behind operational recommendations visible and auditable to the people relying on them in the field.
That is a meaningfully harder problem than throughput, and it is the correct one to prioritize at this stage of TracPlus's maturity. A platform that already reliably ingests a million records a day has largely solved the scale problem; the next-order challenge is building algorithms and interfaces that translate that volume into planning and operational decision support agencies can trust and act on quickly, which is a product and trust problem as much as an engineering one. Enterprises further behind on their own data platforms would do well to resist the urge to chase throughput past the point where it still moves a real decision.
Why this matters beyond wildfire operations
TracPlus is a small company relative to the hyperscalers and enterprise software vendors that usually populate this column, but its engineering challenge, real-time ingestion of massive geospatial and sensor data streams feeding decision support for safety-critical operations, is structurally identical to problems facing utilities managing grid sensors, logistics operators tracking fleets, and any enterprise building operational intelligence on top of IoT-scale data volume. The lessons from how Cheng approaches this transfer well beyond aerial firefighting, which is exactly why a niche operational technology hire like this belongs in a broader enterprise leadership conversation.
The specific discipline worth borrowing is her explicit distinction between raw data transparency and decision transparency. Many enterprise data platforms optimize heavily for the former, dashboards, reports, and queryable data lakes, while leaving the latter, why a system recommended a specific action, opaque even to the operators using it daily. Cheng's stated priority of exposing the reasoning behind decisions, not just the underlying data, is a useful design principle for any team building operational AI or analytics tools where end users need to trust, not just consume, the output.
What we would watch next
The real test of this hire will show up in TracPlus's next platform releases: does the Mission Platform start surfacing explainable recommendations to fire agencies, or does it continue functioning primarily as a high-reliability tracking and reporting layer with decision-making left entirely to human experience, as Cheng described the current state. If the former materializes over the next several quarters, it validates that a payments-and-crisis-communications background transfers cleanly into safety-critical operational AI, a talent pathway most enterprise recruiters are not yet actively mining.
For technology leaders building or evaluating operational intelligence platforms in any high-consequence domain, TracPlus's approach is worth tracking as a proof point independent of its niche market. A billion-record historical dataset and a million-record daily ingestion rate are meaningful engineering achievements on their own, but Cheng's stated ambition, making the reasoning behind decisions visible rather than just the data behind them, is the harder and more valuable problem, and it is the one that will determine whether this hire is judged a success.
.webp)