How AI Agents Are Closing the Execution Gap in Sustainability

Atsuya Tech addresses a persistent operational blind spot faced by enterprises striving for Net-Zero and regulatory compliance. While companies deploy sensors, dashboards, and analytics, execution on the ground often lags. Atsuya solves this by unifying energy, air quality, refrigeration, and facility assets onto a single platform, AOne, that translates sustainability goals into real-time operational action.

Their customers range from quick-commerce companies, where freshness and uptime are mission-critical, to hospitality and commercial real estate players, where indoor air quality, energy efficiency, and carbon footprint directly affect compliance and customer experience. For years, Atsuya relied on IoT systems and predictive AI to detect anomalies. But alerts alone weren’t enough. They wanted to move from detecting problems to preventing them altogether. AI Agents became the natural next step.

Participation in the AI Agents Masterclass aligned with their long-term vision of creating a Digital Facility Manager, Virtual Energy Auditor and Cooling Specialist - agentic systems capable of handling up to 80% of routine compliance and optimization tasks without human intervention.

This led to the creation of the AOne Agent, a multimodal AI Agent. Acting as a 24/7 Asset Controller, the Agent continuously monitors indoor air quality, occupancy, energy loads, and asset health and takes corrective action autonomously.

Building an autonomous agent in such environments came with real challenges. In quick commerce, downtime is unacceptable - a failed freezer can mean instant inventory loss and revenue impact. The Agent could not afford to hallucinate or take unsafe actions. Atsuya implemented strict guardrails and deterministic overrides, hard-coded safety limits that the AI cannot bypass.

Another key insight was the importance of contextual awareness. An energy spike at 9 a.m. may be a normal startup load; the same spike at 9 p.m. likely indicates a leak. Giving the Agent temporal and operational context significantly reduced false positives and improved trust.

Atsuya is currently in the pilot phase with select enterprise customers. Early feedback has been strong—users value that they no longer need to constantly monitor dashboards. Instead, the Agent notifies them after it has resolved minor issues, delivering what many describe as “peace of mind.”

More importantly, the AOne Agent has shifted Atsuya’s positioning from an IoT dashboard provider to an AI-driven operations partner. This evolution is already opening doors to outcome-based contracts, where Atsuya shares directly in the operational and energy savings it delivers.