AI is moving from demos into real-world enterprise applications, but real-time data remains a key challenge.
How can enterprises build a lighter, faster data foundation to deliver real-time data to AI?
On August 20, DolphinDB and EMQ co-hosted a live session on building lightweight data infrastructure for AI applications in industrial and financial scenarios.
New Challenges in Enterprise Digital Transformation: How Can AI Make Better Use of Data?
As enterprises move from traditional data analytics to AI-powered queries, predictive analytics, intelligent decision-making, and AI Agents, the challenge is no longer just having data—it is making data accessible to AI in real time.
In a typical enterprise architecture, data flows through multiple layers:
Business Systems → Messaging → Stream Processing → ETL → Storage → Analytics → AI / BI
This fragmented architecture increases maintenance costs and implementation time, while making it difficult for AI to access timely, reliable data.
For example, when an operations team asks, “Which devices showed abnormal temperatures over the past 24 hours?”, an enterprise Agent needs to understand the relevant data sources, fields, units, thresholds, and business context before it can deliver a reliable answer.
The key to bringing AI into production is therefore not just a more powerful model, but a data foundation that can deliver the right data to AI in real time.
DolphinDB + FlowMQ: Building a Lightweight Real-Time Data Foundation
To simplify complex enterprise data architectures, the session introduced the joint solution of DolphinDB + FlowMQ.
FlowMQ handles data ingestion and real-time streaming, DolphinDB handles storage and computation, and AI turns the data into business value.
FlowMQ connects data from sources such as PLCs, MES, ERP, market data, trading systems, and logs, while DolphinDB provides real-time storage, analytics, and computation.
Together, they simplify the data pipeline to:
Data Sources → FlowMQ → DolphinDB → AI / BI / Dashboard
A shorter data pipeline means lower latency, simpler infrastructure, and faster access to real-time data for AI-powered decision-making.
DolphinDB: A Lightweight Data Foundation to Escape the “Data Infrastructure Trap”
According to customer research, 80% of enterprise AI teams’ time is spent on data infrastructure—connecting data pipelines, aligning metadata, cleaning data, and managing complex technology stacks. Only 20% is left for business model optimization and value creation.
To address this challenge, DolphinDB provides a lightweight, high-efficiency data foundation with three core capabilities:
- Break down data silos: Multimodal storage engines and external tables provide unified access to diverse data sources, including MySQL, SQL Server, HBase, S3, and Parquet.
- Native advanced analytics: With 2,000+ built-in analytical functions, DolphinDB supports financial and IoT use cases, from signal processing to machine learning. Complex analytics can be performed directly in SQL, with performance up to dozens of times faster than Python or Spark + Hive.
- All-in-one platform: DolphinDB integrates storage, computation, stream processing, and AI analytics into one platform, eliminating fragmented architectures and reducing development cycles from months to weeks.
DolphinX: Making Enterprise-Grade Agents Ready to Use
As AI Agents move into production, enterprises need them to understand business data and safely interact with enterprise systems. DolphinDB introduced DolphinX, an enterprise-grade AI development and governance platform, to address this challenge.
DolphinX helps Agents understand data definitions, field meanings, and business rules, while features such as automatic context management, long-term memory, Skills, and MCP enable them to continuously adapt to specific business needs.
As Liang Lin, Solution Director at DolphinDB, noted, enterprises need more than generic LLM capabilities—they need AI that understands their data, knowledge, and business processes. DolphinX connects these elements, helping AI move beyond answering questions to truly understanding and serving the business.
EMQ FlowMQ: A Cloud-Native Platform for Real-Time Data
If DolphinDB makes data easy to store and analyze, EMQ focuses on making data easy to connect and stream in real time.
Liang Chen, Director of Financial Industry Solutions at EMQ, noted that AI competition is ultimately about who can access more real-time data. Yet many enterprises still face bloated architectures and the operational complexity of Kafka and Flink.
FlowMQ: One Engine, Three Messaging Paradigms, Four Protocols
EMQ introduced FlowMQ, a unified real-time messaging platform designed for the AI era:
- Multi-protocol support: Native support for MQTT, Kafka, AMQP, and NATS, enabling direct integration without complex bridging or conversion.
- Simplified architecture: One engine supports publish/subscribe, queue, and stream processing, reducing the need for multiple messaging systems and potentially cutting operational workload by over 50%.
- High performance, lower cost: Delivers up to 1 GB/s throughput, 2 million messages/s, and P50 latency below 1 ms for long connections, while cloud-native storage separation helps reduce TCO.
From Industrial to Financial Applications: Real-Time Data Foundations Across Industries
Real-time data infrastructure supports a wide range of industries.
In financial services, FlowMQ handles real-time data from markets, trading, and risk management, while DolphinDB provides stream processing, factor computation, and CEP for quantitative analysis and market data distribution.
In industrial settings, FlowMQ connects to PLCs, SCADA, MES, ERP, and IoT systems, while DolphinDB enables time-series analysis, equipment monitoring, predictive maintenance, AI-powered quality inspection, and energy analysis.
Together, DolphinDB and EMQ connect real-time data from devices and business systems with high-performance computing, turning it into data ready for AI and business applications.
In the AI Era, Enterprises Need to Redefine Their Data Foundation
Bringing AI into production requires more than a powerful LLM. Enterprises need a data foundation that can deliver real-time data to AI, turn business knowledge into reusable capabilities, and connect AI with real-world workflows.
FlowMQ enables real-time data ingestion and streaming, while DolphinDB transforms that data into analyzable and reusable data assets. Together, they provide a unified real-time data foundation for enterprise AI.
Lighter. More real-time. Simpler.
This is the direction DolphinDB and EMQ are working toward to help enterprises turn AI into real business value.
For the live session slides and demo materials, please contact us at info@dolphindb.com.