Code Moment | DeepWise Closes Hundreds of Millions of RMB in Series A and A+ Rounds, Existing Shareholder Source Code Capital Continues to Increase Stake
Building Next-Generation Deep Energy-Saving Intelligent Control Products and Solutions
DeepWise recently raised several hundred million RMB in Series A and A+ funding, jointly led by Shenzhen Capital Group and China Merchants Capital, with strategic investment from Guangyuan Jinpan Industrial Fund. Existing investors HSG and Source Code Capital continued to increase their stakes. This round marks DeepWise's fourth financing round following investments from Tencent, HSG, Source Code Capital, and Inovance. The proceeds will primarily fund new product R&D and business expansion.
Founded in 2018, DeepWise (DeepCtrls) is a deep energy efficiency and digital intelligence innovation service provider focused on empowering industrial and building energy sectors to achieve deep energy savings in electromechanical systems and digital transformation. The founding team hails from Tsinghua University and Lawrence Berkeley National Laboratory in the United States, bringing two decades of experience in energy efficiency.
As China's dual-carbon strategy takes concrete shape, demand for energy conservation, carbon reduction, and digital transformation across industries has grown increasingly urgent. In industrial and building energy systems, HVAC and compressed air systems often account for 30% to 50% of total energy consumption. Yet improving the operational efficiency of these systems still largely relies on three approaches: equipment-based energy savings (essentially Energy Efficiency 1.0), expert experience-driven savings (2.0), and purely AI-powered passive optimization (3.0) — resulting in hundreds of billions of kilowatt-hours wasted annually in China.
Is there a new technical approach that can approach the theoretical limits of operational energy savings?
"A new era demands a new generation of energy efficiency platforms that combine deep domain expertise with data-driven approaches," said Dr. Hui Li, founder of DeepWise. Unlike "equipment-based savings," "expert experience savings," or "AI groping in the dark passive savings," the company has developed an industry-leading fourth-generation "mechanism framework + data-driven" high-precision modeling and optimization technology. For electromechanical systems like HVAC and compressed air, this shifts the paradigm from passive to active, from local to global optimization — in real time, searching through hundreds of thousands of possible operational routes to identify the most efficient path and approach the system's energy-saving limit. This "fourth-generation" optimization technology provides DeepWise with the technical foundation and competitive edge to tackle demanding scenarios in advanced industrial manufacturing and data centers characterized by "high complexity, high precision, and high requirements."
Since its founding, the company has iterated its technology, refined its products, and pushed commercialization forward. How has this company — which now covers numerous leading clients in the deep energy efficiency track — achieved this?
"Saving Beyond Savings": Approaching the Energy Efficiency Limit Next-Generation Technology Empowering the Industry
Energy cost reduction is a persistent need and pain point across industrial facilities, IDC, and numerous other scenarios. The challenge for today's next-generation energy efficiency control platforms is how to achieve "savings beyond savings" on top of existing energy-saving control systems, fully tapping into the system's energy-saving potential. Typically, the difficulty in optimization lies in the coupled performance interactions between various equipment and systems — with hundreds of thousands of possible operational combinations to meet any given end-load demand. Expert rule-based approaches cannot find optimal solutions in real time, while pure AI algorithms trained on historical operational data lack physical meaning in their models. Their optimization results are constrained by the quantity of historical data and the accumulated tuning experience of veteran operators, making system-wide optimization and substantial energy savings elusive.
DeepWise's solution for moving from savings to "savings beyond savings": First, build high-precision simulation models and predictions for HVAC equipment. While guaranteeing end-load requirements, the system real-time traverses and calculates the energy consumption corresponding to hundreds of thousands of system-level operational condition combinations in a simulated environment, thereby identifying the optimal control parameters that minimize total system energy consumption, and implementing closed-loop control and global proactive optimization at the system level. Currently, the error between simulated performance predictions and actual values is generally kept within 3% — meaning the system can in real time approach the operational energy-saving limit within a 3% margin.
From the client's perspective, energy savings are quantifiable and perceptible. Through a one-click switching function, clients can compare pre-optimization and post-optimization data within five minutes, enabling real-time multi-dimensional verification of energy savings. Based on statistics from hundreds of deployed projects, DeepWise's energy efficiency platform can achieve at least an additional 10% improvement in energy savings rate on top of market-leading expert experience-based or AI-based energy savings.
Agile delivery has also been key to DeepWise's business breakthroughs. In a Tencent IDC project, one or two engineers completed system configuration and debugging, going from initial engagement to go-live in two weeks — with zero downtime or operational impact to on-site production. Li told 36Kr Carbon: "Through rapid, lightweight deployment and delivery, we already reduced on-site configuration time from two weeks to one week last year. This year's target is to get it under three days." Beyond deep energy savings, the high-precision simulation models also enable fault early warning, predictive equipment diagnostics and maintenance, and safety protection and interlocking — enhancing operational safety assurance.
Significant, Quantifiable, Verifiable Savings Driving Customer Repurchases
The value delivered by "savings beyond savings" is a critical driver of market demand. And the key metric of value is quantifiable, verifiable return data. Indeed, tangible economic returns have brought DeepWise continuous add-on purchases and repurchases from clients, enabling networked market expansion. In one deep energy efficiency and digital intelligence factory project for a major new energy battery industry leader, DeepWise started with just a single pilot building. Based on the actual annual savings of 7.5 million kWh from that first pilot, the client gradually connected other buildings in the park to the energy efficiency and intelligent control platform — achieving deep energy efficiency control and intelligent O&M for nearly 80 electromechanical energy systems across the park, with over 80,000 monitoring points. Annual HVAC system energy efficiency improved by over 21%, annual operating costs were reduced by tens of millions of RMB, and average O&M response time was cut by over 70%. Meanwhile, production environment control precision and safety assurance were further enhanced.
To date, through empowering ecosystem partners, DeepWise has delivered deep energy efficiency and digital intelligence systems for hundreds of client projects. These include advanced manufacturing factories and data centers such as CATL, BOE, the National Supercomputing Center, Tencent Data Center, Great Wall Motor, Guobo Electronics, Huatian Technology, GE, and Beijing Capital Agribusiness Group, as well as rail transit, energy stations, top-tier hospitals, and landmark commercial buildings.
Li stated that DeepWise will continue strengthening its product and technology moat, solidifying its technical leadership in industrial and building energy efficiency, further developing cross-industry applications from existing electromechanical energy systems to distributed energy and integrated energy systems, and expanding overseas. "Our vision is to build world-class deep energy efficiency intelligent control products and solutions, empower the industrial and building energy efficiency industry, and become a global leader in energy efficiency intelligent control."
