Executive Summary
A smart-building platform that makes commercial buildings more energy-efficient and secure by adding AI on top of the infrastructure and low-cost sensors they already have, instead of making them rip out and replace it. I owned the venture plan solo: market sizing, product vision, technical architecture, business model, MVP scope, and go-to-market roadmap.
Built over six months as my Kellogg PM certificate capstone (2024); presented and awarded full marks, not taken to market.
Challenge
Buildings account for 40% of US energy use, and commercial buildings waste about 30% of the energy they consume (public research). Existing "smart building" technology doesn't close that gap: it can't measure usage, take automated action, or apply modern AI/ML, and what does exist is often too expensive for most buildings to adopt.
Safety has a similar gap. Tenants want entry points that lock automatically when they leave, alerts to authorities on real threats, and an easy way to let a guest in without the hassle of security.
Approach
Market sizing: instead of sizing the broadest possible global market, I focused on the market I could realistically win: C40 Climate Pact cities in North America (LA, SF, Toronto, Vancouver, NYC), climate-forward cities where building owners are already inclined to adopt this technology.
Build vs. buy on hardware: to keep cost and the barrier to adoption low, I built on what a building already had (WiFi, cameras) and added low-cost IoT sensors only where needed, instead of designing custom hardware.
Business model: that choice shaped the business model: a software subscription plus leased hardware, instead of an upfront hardware sale. Leasing lowered cost and commitment for buildings that already had most of the infrastructure.
Execution
The capstone scoped an MVP as the proof of concept: an analytics dashboard for building managers, a tenant mobile app, and monitoring through existing WiFi and IoT sensors. The fuller roadmap followed: an on-site edge controller, camera integration, and AI/ML add-ons like predictive maintenance and facial recognition.
Outcome
Delivered the full venture plan, presented it to a broader audience, and received full marks for the project and the course. Traction figures in the deck are projections.
Reflection
If I rebuilt this today, I'd push the "use what's already there" principle even further: designing the whole platform around off-the-shelf signals (WiFi, existing IoT devices, thermostats, cameras) and cheap sensors plus AI, with effectively no proprietary hardware. That would make it easier for a building to adopt, without a heavy hardware investment even as a supplement.