I'm Ezekiel (Zeke) Pari, a Brisbane-based product and systems builder and Computer Science student. I co-founded MESR, where we're building AI-assisted condition assessments for infrastructure.
Right now, the thing I care most about is MESR. We're building a configurable infrastructure inspection app that turns photos and video into AI-assisted condition assessments. Field teams capture through the method that fits the site; civil and asset engineers review the condition, supporting imagery, and uncertainty before moving the assessment into existing inspection and maintenance workflows.
My work spans product and systems across that workflow, from flexible capture and AI analysis to engineering review and operational handoff.
Outside Australia? • Australian Qualifications Framework
6.75 GPA • Minor in Cloud Computing • Graduating 2026
These are not here to prove I can code. They show a pattern: find an operational problem, turn it into a useful product or repeatable system, and learn from real users. MESR is where I'm compounding that pattern now.
RoGold put my work inside a consumer product serving more than 800,000 users. My focus was helping turn a growing collection of features into a more coherent commercial product: clearer onboarding, multilingual reach, subscriptions, live integrations, and interactive demonstrations.
I also helped unify the free and paid product architecture while reducing core code by roughly 70% without removing functionality. It taught me that at scale, migration, reliability, revenue, and product coherence matter as much as shipping the next feature.
SmartCall AI started with a simple commercial failure: when a business misses a call, it can lose the booking, lead, or customer attached to it. I built a voice product that could answer, understand intent, capture the opportunity, book appointments, and hand a useful summary back to the team.
The founder work was translating that promise into an operation: configurable agents, escalation paths, role-based dashboards, CRM-ready records, and infrastructure that could handle demand in bursts. The technology served the business outcome, not the other way around.
Prompt Solutions began with ordinary website work, then asked a better business question: how can good delivery become repeatable instead of starting from zero for every client? I turned recurring work around content, assets, portfolios, approvals, and publishing into shared product workflows.
The result was a productised service with client dashboards, media handling, before-and-after tools, portfolio systems, controlled feature rollout, and subscription billing. It is the clearest example of me trying to build an operating system around a service, rather than simply selling pages.
RoLinker turned repetitive identity and permissions work across Roblox and Discord communities into a self-service product. It gave users control of linked accounts while giving operators the divisions, dashboards, automation, and developer API needed to run larger communities without scaling administration at the same rate.
WorldMC was more than a game website. It was the information and market layer for a player-run world: people needed to understand players, towns, nations, territory, and prices before they could act inside the economy. I built the public product and the underlying API around those decisions.
Shockwaves Radio combined a public media product with a staffed operation. I built the system behind the station—live streaming, playlists, schedules, roles, and partner integrations—so the team could run programming consistently rather than depend on ad hoc coordination. It taught me to design for operators and handoffs, not only audiences.
Thanks for reading.
me@zekepari.dev