By Ayten Abdelrazek – September 10, 2025
“AI is moving fast. The real edge? Using your data, your tools, your way.” — Heather VanCura, Oracle, TechCrunch 2025
At TechCrunch 2025, Oracle took the stage with a clear message: AI doesn’t have to be complicated, and you shouldn’t need to duct-tape half a dozen services together just to get value from your data.
In a packed session led by Heather VanCura, a Vice President at Oracle, and Sandeep Agrawal, one of the original architects of MySQL HeatWave, we got a glimpse of a radically simplified, developer-first AI future.
Oracle’s Vision: AI That Lives Where Your Data Lives
If you’ve ever built anything with AI, you probably know the chaos: data in one place, models in another, analytics in a third… and endless pipeline pain.
Heather framed it perfectly: “The AI horizon requires a foundation from which we can ask and answer questions that haven’t even been thought of yet.” That foundation? Your data, in one place, powered by Oracle’s full-stack AI infrastructure.
Here’s how Oracle is changing the game:
- Data-first AI: Enterprise-grade data meets both small and large LLMs.
- End-to-end stack: From OS to cloud infra (OCI, Java, MySQL, Linux, and more).
- Ultimate flexibility: Choose your cloud, your models, and your language (especially Java!).
AI & You: A Developer’s Paradise
This isn’t just another AI framework. Oracle is bringing AI right into the MySQL developer experience:
- Java SDKs + VS Code plugins
- In-database ML & GenAI
- REST APIs, SQL, JDBC—all supported
- No GPU? No problem. It still works!
Introducing MySQL HeatWave: One Database. Every Capability
At the heart of Oracle’s announcement was MySQL HeatWave, a revolutionary cloud service that eliminates the traditional need to move or replicate data for AI, analytics, or transactions.
Sandeep Agrawal brought the fire to TechCrunch—literally. His demo showed how MySQL HeatWave can power a real-time GenAI medical app in just a few REST calls.
What is HeatWave?
- Multi-Cloud Database Service: HeatWave is Oracle’s multi-cloud database platform, designed for maximum flexibility and performance.
- Unified Data & AI Stack: HeatWave seamlessly integrates transaction processing (OLTP), analytics, machine learning, and generative AI within a single database engine.
- Real-Time, Low-Latency Insights: New data is available instantly for AI and analytics, eliminating the need for time-consuming data syncing or batch processing.
- Flexible AI Integration: Users have the option to choose secure in-database large language models (LLMs) or connect to external AI providers such as LLaMA, Cohere, and OpenAI.
- Developer-Friendly Platform: HeatWave supports familiar tools, including REST APIs and full Java compatibility, empowering developers to build robust, AI-powered applications efficiently.
From Patient Notes to Production AI: A Developer-Centric Demo
During the session, attendees witnessed an end-to-end demonstration of what it looks like to build a secure, generative AI-powered healthcare application—without leaving the database or compromising privacy.
Imagine:
- A doctor recording notes during patient visits, stored directly in MySQL.
- Using a simple REST API, those notes are summarized in real time with generative AI.
- A second API call anonymizes personally identifiable information (PII) and formats the content for email—all processed securely within the HeatWave database, without exporting data to third-party tools.
This seamless experience illustrates Oracle’s commitment to removing complexity so developers can focus on innovation—not complications.
Thought question: What would it mean for your app if every new row of data was instantly queryable by AI?
ML + GenAI: Better Together
One of the most impactful parts of the session was Oracle’s perspective on the synergy between traditional machine learning and generative AI.
Sandeep described several real-world customer use cases where this combination is driving tangible results.
Real World Application Solutions:
- IT Log Analysis: Anomaly detection models flag unusual patterns in server logs, and LLMs then summarize root causes—cutting investigation time dramatically.
- Financial Fraud Detection: ML identifies suspicious transactions (e.g., mule accounts), and GenAI explains the behavior behind those patterns.
- Food Delivery Personalization: Recommendation systems identify preferred restaurants, and GenAI suggests customized menu items by parsing thousands of PDF menus.
These are not hypothetical examples—they are in-production deployments by banks, delivery services, and tech companies solving real-world challenges at scale.
Bonus: These run at scale—up to 512 nodes and 8,000 cores. Enterprise-level horsepower, no duct tape required.
Built for Scale, Security, and Simplicity
Oracle’s vision is grounded in practical, scalable architecture:
- Secure by Design: In-database AI keeps your data within Oracle Cloud, minimizing exposure and meeting compliance standards.
- Massive Scale: HeatWave can handle up to 512 nodes and 8,000 cores—processing hundreds of millions of records, from object stores to relational tables.
- Open and Flexible: Java SDKs, REST interfaces, VS Code plugins, and native connectors make integration seamless for modern developers.
Calling All Builders: Get Started Free
Heather and Sandeep made one thing crystal clear at TechCrunch:
“You don’t need a data science army. Just your data, your code, and MySQL HeatWave.”
Here’s how to jump in today:
- Try the always-free HeatWave tier
→ oracle.com/mysql/heatwave - Use $300 free cloud credits to scale up
- Install the HeatWave VS Code plugin
- Explore AI & ML notebooks on GitHub
- Join the conversation at Oracle for Developers
The Future is Oracle + Data + AI
We’re entering a new era. The AI winners won’t be those with just the biggest models—they’ll be the ones who turn their data into answers, fast, using tools their teams already love.
So whether you’re:
- A dev ready to embed AI into your app
- A data engineer tired of fragmented pipelines
- Or a founder exploring what GenAI means for your product
Oracle’s message from TechCrunch is this: It’s time to simplify, scale, and build.
Ready to create the next generation of AI-powered apps?
And if you’re lucky enough to catch Heather VanCura or Sandeep at an Oracle event near you, don’t miss it. They’re not just talking AI—they’re building the future of it.
