MDBS Builds a Comprehensive FinTech Ecosystem with OWL AI, Lobster Manor, and Speedy FPGA

On August 21, MDBS Digital Technology hosted the “Brain of Tomorrow: Financial AI Solutions × Intelligent Data Centers” seminar. Through real-world use cases and live demonstrations of core applications including OWL AI and Lobster Manor, MDBS Chairman Simon Chang explained how the company transforms structured data into high-performance API services—enabling AI to evolve from a basic question-and-answer tool into a professional platform for market research and financial services.

OWL AI Integrates Market Data, Financial Statements, and Strategy Components

Traditional generative AI models often rely on web searches and data scraping when responding to financial queries. However, financial data is highly time-sensitive, continuous, and structured. Missing monthly records, unadjusted prices, or inconsistent formats can directly compromise the accuracy of analytical results.

OWL AI integrates a broad range of financial market data, including real-time quotes, transaction details, daily, weekly, and monthly candlestick charts, intraday charts across different time intervals, and technical indicators such as moving averages, KD, RSI, MACD, and Bollinger Bands.

Its fundamental data coverage includes revenue, earnings per share, profitability, balance sheets, cash flow, dividend policies, and financial ratios.

During the event, a comparative test was conducted to calculate the adjusted returns of a 15-year dollar-cost averaging strategy for the Yuanta/P-shares Taiwan Top 50 ETF (0050).

The results showed that when AI relied on web scraping and its own reasoning, it consumed 766,301 tokens but retrieved only 132 months of data, representing approximately 73% coverage. In contrast, by accessing structured data through OWL AI, the model consumed only 45,921 tokens and retrieved complete and accurate data covering all 180 months.

The test demonstrated that structured data required only 6% of the tokens consumed by conventional web scraping, while reducing API costs to approximately one-tenth and producing results that fully matched the verified data.

These findings show that direct access to a high-quality database can minimize inefficient searches, significantly reduce computing costs, and prevent hallucinations caused by incomplete data.

Lobster Manor: Turning Structured Data into Real-World Applications

Beyond data services, Simon Chang, Chairman of MDBS Digital Technology, publicly unveiled Lobster Manor for the first time, demonstrating how financial data can be transformed into a practical end-user application.

Positioned as an intelligent stock analysis tool, Lobster Manor integrates individual stock prices, financial indicators, and AI-generated analysis within a visual interface, allowing users to assess market conditions more intuitively.

The demonstration highlighted how AI can accurately retrieve information from financial databases, organize data in real time according to user instructions, and present professional analysis through accessible text and visuals. This completes the service chain from foundational data provided by OWL AI to the front-end application delivered through Lobster Manor.

Speedy FPGA Establishes the Foundation for Ultra-Low-Latency Trading

AI applications depend on accurate data, while real-time financial services require an exceptionally fast underlying infrastructure.

During the event, Simon Chang presented the Speedy FPGA financial trading architecture. Powered by hardware acceleration technology, the solution enables high-speed data processing and meets the financial market’s stringent requirements for ultra-low latency.

Speedy FPGA is planned in two configurations:
·Professional Edition: Approximately 1.8 microseconds of latency
·Ultra-Speed Edition: Approximately 1.5 microseconds of latency

The system can be flexibly configured according to the number of connections, FIX sessions, and financial products.

Speedy FPGA complements OWL AI and Lobster Manor by addressing three distinct layers of financial technology: underlying trading performance, structured data delivery, and advanced analytical applications. Together, these solutions establish MDBS’s comprehensive technological foundation and competitive advantage in the FinTech sector.

Leveraging Financial Expertise to Accelerate Real-World AI Adoption

The value of financial AI lies in its ability to apply accurate data to real-world challenges. MDBS has extensive experience in financial trading and information systems, with a deep understanding of the industry’s rigorous requirements for real-time performance, low latency, and cybersecurity./p>

Looking ahead, MDBS will continue to deepen the integration of data and AI technologies, helping financial institutions shorten the journey from proof of concept to full-scale deployment while delivering reliable, scalable, and high-value financial solutions.