Zync: Leading an AI/video SaaS from interface to infrastructure
I joined Zync as a frontend developer and gradually took ownership across backend systems and cloud infrastructure, helping shape five products across the stack.
AI/video infrastructure - Make video processing reliable at a cost you can understand.
Video products often work well in a demo, then meet expensive rendering, uneven demand, retries, and slow failure diagnosis in production. I help teams trace those costs and delays through the processing path and make focused improvements.
Where the work starts - Find the costly or unreliable step first.
The useful starting point is a real production path: what starts a job, where it waits, what it computes, how it fails, and how the output reaches a user. That makes the next engineering change specific and measurable.
- Rendering economics. Attribute cloud spend to the jobs and resources that create it. Look for idle capacity, repeated work, or a processing step whose cost rises faster than output.
- Queues and throughput. Check how jobs enter the system, wait, retry, and complete under normal and peak demand. Improve the bottleneck that limits useful output.
- Production recovery. Make failures visible and actionable. Clarify ownership, alerts, and recovery paths so an incident can be diagnosed and resolved.
Relevant experience at Zync
90%lower cloud compute cost
I led queue-based rendering work across Zync’s AI/video stack. The resulting system supported 200 renders per day, with peaks around 500.
Read the Zync case studyEngagement path - A practical route from diagnosis to improvement.
The first deliverable can be a technical assessment or a bounded production change. We choose the shape after looking at the system, the available measurements, and the cost of the problem.
- 01 / Baseline
Map the job lifecycle
Review the processing path, timings, failure modes, and cost signals to establish what needs to change.
- 02 / Change
Fix the limiting step
Implement a targeted change to the queue, worker, media processing, or cloud setup as the evidence directs.
- 03 / Verify
Measure in production
Compare cost, throughput, and reliability against the baseline and leave the team with a clearer operating picture.
Selected work - Product delivery connected to production systems.
At Zync, my role grew from front-end development into backend, AI/video workflows, and cloud operations across five product areas. See how product delivery, rendering, and reliability connected in one platform.
Where does your video pipeline slow down?
Send the current processing path, the volume you handle, and the cost or reliability issue you see. We can decide whether an assessment or a focused change is the right first step.
Location and availability
- Global · Flexible
Remote · global teamsOpen to conversations nowUp to 30–40 hours/week