Bonsai DevTool \Canada

Step 2 - Validation: Live Deployment and Savings Verification (Months 4-6). Switch from shadow mode to live control in the pilot building. Run the RL policy for 90 days, continuously learning and adapting to occupancy patterns, weather, and equipment performance. Use a rigorous measurement and verification protocol (IPMVP standards) to calculate actual energy savings compared to a baseline. Publish a case study with verified savings data, ROI calculations, and customer testimonials. Use this case study to close 3-5 additional customers in the same vertical (either commercial real estate or data centers, not both initially). Charge customers 30% of verified annual energy savings as a recurring fee.

Step 3 - Growth: Productize Deployment and Scale Sales (Months 7-12). Build a self-serve digital twin creation tool that ingests building blueprints, equipment specs, and historical sensor data to auto-generate simulation environments. This reduces deployment time from weeks to days. Develop a library of pre-trained RL policies for common HVAC configurations (rooftop units, chillers, air handlers) that can be fine-tuned per building. Hire a sales team focused on commercial real estate property managers and data center operators. Offer a risk-free pilot program: no upfront cost, we only get paid if we deliver verified savings. Target 20-30 buildings under management by end of year one. Build integrations with major BMS vendors (Johnson Controls, Siemens, Honeywell) to streamline deployment.

Step 4 - Moat: Expand to Adjacent Verticals and Build Data Network Effects (Months 13-24). Once HVAC control is proven, expand to adjacent building systems: lighting optimization, demand response, predictive maintenance for chillers and boilers. Aggregate anonymized data across all managed buildings to train better RL policies—buildings with similar characteristics can bootstrap from each other's learned behaviors. This creates a data moat: the more buildings we manage, the better our policies perform, the faster we can deploy new customers. Explore strategic partnerships with utilities (who will pay us for demand response capabilities) and ESG reporting platforms (who need verified carbon reduction data). Build an API that allows customers to integrate our optimization engine into their own energy management dashboards. Long-term vision: become the operating system for intelligent buildings, with HVAC as the wedge and RL as the core technology.