Deinland Solar accelerates app development with Oracle Autonomous Database 23ai

Deinland Solar is a German solar energy startup, focused on getting solar-power installations on commercial building rooftops and suitable open spaces. The company used Oracle APEX running on Oracle Autonomous Database 23ai to build its ENIGNUM mapping and analysis application to help salespeople identify opportunities for solar panel installations across Germany. By integrating spatial, energy, and government-supplied property value and ownership data, the app lets salespeople view detailed maps, zoom into specific buildings, and assess solar potential by analyzing roof characteristics, energy consumption, and potential cost savings. The AI and Oracle Spatial features in Autonomous Database 23ai support quick data refreshes and queries on large spatial data sets for faster client outreach and proposal generation.

Deinland Solar built its ENIGNUM application with Oracle APEX, and chose Oracle Autonomous Database 23ai on Oracle Cloud Infrastructure (OCI) to handle the real-time, AI-powered processing of complex data in many formats, including detailed property and land registry data, specific building and roof characteristics, and precise geographic location and structural details. Using the Oracle Spatial features in Autonomous Database 23ai, the company could rapidly process the massive volumes of geospatial and engineering data used within ENIGNUM. Also, the AI Vector Search and Select AI features in Autonomous Database 23ai interested the company, which plans to connect large language models to internal data sources as a resource for business insights. Deinland wants to use retrieval augmented generation (RAG) to inform LLMs so they can access company data while keeping that internal data private.

Deinland Solar gained the scalability it needed for growth using Oracle Autonomous Database 23ai on OCI. The company can handle huge data volumes with high-speed performance and nearly zero maintenance. Deinland developed the ENIGNUM application using Oracle APEX, a low-code development platform that’s included with Autonomous Database 23ai. Oracle APEX allowed the company to easily add a generative AI-powered conversational interface to its application. Autonomous Database 23ai simplified database administration and handled approximately 95% of DBA tasks automatically. This approach let the company’s one-person development department build a sophisticated app much faster and with less overhead compared to traditional approaches.

Using ENIGNUM running on Autonomous Database 23ai, sales teams can now identify strong solar power opportunities and make recommendations to companies. ENIGNUM has to manage complex data across multiple dimensions: tracking new solar panel installations, monitoring ground-based electrical infrastructure such as overhead power lines and transformer stations, analyzing electrical grid connections, and maintaining detailed records of customers’ energy production and asset performance. The app uses government APIs for access to cadastral information and spatial data collection. All this data is needed to ensure salespeople are working with accurate and official data when they make recommendations. Autonomous Database 23ai’s Oracle Spatial capabilities gave Deinland Solar a centralized data source, which bolstered its ability to process and use myriad data sets quickly, including 150 million spatial objects within the database. And because spatial data changes over time and requires updating, Deinland Solar easily integrated new data as needed, sometimes refreshing as many as 30 million new spatial objects daily.

The startup is exploring AI Vector Search for more refined roof shape classification and analysis by converting spatial and visual data into vector embeddings. Deinland Solar is also interested in using OCI Vision to map and categorize solar panel locations, ground infrastructure, and electrical lines via AI-powered visual recognition. These capabilities would give ENIGNUM the ability to analyze and extract information from visual data, such as satellite images of buildings, to determine the precise shape, size, and orientation of a roof.

Additionally, the company is looking into the use of Autonomous Database Select AI in conjunction with RAG to create a new user interface to let people ask chatbot-style questions of the company’s data sources. The tool would help salespeople more easily access information about roof characteristics and solar potential.