Tendai Ellison Mudavanhu
I’m an engineer in Melbourne working across network planning, automation, data platforms and AI-assisted workflows. Most of my career has been in telecommunications; lately I have been building small agent and LLM systems around practical workflows I understand.
What I’m building
Current builds, product ideas and work notes.
Gombwe is my personal assistant and tool-use runtime. I reach it through Discord and connect it to local workflows on my Mac and home network: groceries, school events, meal planning, parental controls, home-network checks and recurring household admin. It is also where I test patterns for tools, permissions, logs and review points.
AgentsForm is my main AI website. The idea is that as agentic AI becomes more widely used, agent-facing surfaces will matter: places where agents can discover instructions, forms, APIs and skills needed to complete tasks for people and businesses.
AI Receptionist is a smaller build for trades and service businesses: missed-call recovery, customer intake, urgency detection, booking handoff, transcript summaries and follow-up actions.
TradeProcurement is about tender and RFP workflows: reading documents, extracting requirements, comparing suppliers and drafting response material while the commercial decision stays with a person.
Alongside that I’m working through NVIDIA AI-infrastructure certifications: the associate completed, with the professional tracks underway.
How I got here
Fifteen years of engineering work, mostly in telecom networks, performance, automation and data-driven planning. The AI work sits on top of that background rather than replacing it.
Network planning manager at Metcash, across Mitre 10 and TTHG, working on Meridian: a decision-support platform on Microsoft Fabric and AWS for network strategy, performance diagnostics, market analysis and planning decisions.
Senior network engineer and solutions architect at nbn. I worked on fixed wireless engineering, design automation, telemetry, configuration, capacity and service-quality datasets. One project replaced long manual RF/network design workflows with Python-generated engineering artefacts and configuration scripts. Another developed and validated mmWave 5G line-of-sight prediction at scale, checked against measured signal and speed before feeding nbn’s speed engine.
Network engineer at Ericsson, working on performance, troubleshooting and optimisation across infrastructure projects, using KPI and operational data to isolate faults and validate changes.
A short run at Telstra designing metro Ethernet and fibre transport.
Before Australia, I worked across Africa, Europe and China with Astellia, Huawei and Econet: network analytics, call tracing, KPI extraction, RAN optimisation, field support and capacity/performance improvement.
Studied electronic engineering at NUST, back home in Zimbabwe, with a year in the middle on an embedded-systems and FPGA internship at BFH in Bern.
Somewhere along the way I also did a graduate certificate in business at Melbourne Business School.