Introducing 
Express — Go from prompt to pipeline in minutes.
How years of solving data variety led us to create a conversational data engineering platform
When I founded Nexla, I was obsessed with a problem that held back great ideas: data variety. Not data volume or velocity that got all the headlines back in 2016, but variety. The messy, sprawling complexity of enterprise data with hundreds of systems and ever increasing, each with their own formats, schemas, and quirks, holding critical data in silos.
We saw analysts waiting days for data engineers, and data scientists spending 80% of their time wrangling data instead of building models.The bottleneck was access to ready to use data – right data, right format, right system, on-time.
So we built Nexla with a clear north star: serve the data user and make data accessible to them, without forcing them to become integration experts first.
Over the years, we built something powerful. Our platform handled connectors to hundreds of data sources. We created parsers that understood data structures automatically. We invented Nexsets – virtual data products that made messy data clean, accessible, shareable, and trustable.
But we also built something deeper: a knowledge layer. Every connection taught our system something new. Every transformation added context. Every schema we encountered expanded our understanding. We were accumulating metadata intelligence – the AI that knows not just where data lives, but what it means, how it relates, and how it should be used.
Companies like DoorDash, LinkedIn, Autodesk, and LiveRamp trusted us to handle over 1 trillion records every month. We had gotten good at solving the data variety problem.
Then we watched something remarkable happen in software engineering. As code generation in LLM s started to get better and better, software engineering began to change. Tools like Cursor didn’t just assist developers – they transformed how people thought about building software. Lovable, V0, and Bolt made it even easier, as anyone with an idea could describe what they wanted and watch the application get created. The barrier between intent and implementation collapsed.
I remember the exact moment it clicked. I was prototyping mockups or a feature idea, and then instead I decided to simply explain my idea to Lovable. Natural language → working code. The AI understood context, suggested improvements, even caught errors in its own code and fixed it.
I thought: Why are we still making people click and code to move data around? Why isn’t data engineering conversational?
It was time to make Data Engineering easy and Fast. That answer became Express.dev. All our data engineering expertise, converted to knowledge and wrapped into a conversational interface. Under the hood we had something unique: years of accumulated knowledge about data systems, schemas, transformations, and integrations. We had battle-tested connectors. We had a metadata framework that understood relationships between data sources, data models, transformations. We had enterprise-grade security and governance built in.
Now you could simply say:
“Pull customer data from Salesforce and combine it with website analytics from Google and create a data product.”
And Express will understand your intent. Figure out where the data lives. Connect to it or even create a Connector.Transform it. Prepare it for use. In minutes.
Not a prototype. Production-ready data pipelines from a prompt.
Express isn’t ChatGPT with data connectors. It’s an agentic AI system infused with everything Nexla learned serving enterprises at scale.
Here’s what makes it different:
Express doesn’t just process your request – it creates the exact interface you need for your task. Need to map fields? It generates an interactive mapping view. Need to preview transformations? It builds a data preview. The UI adapts to you.
When you describe what you need, Express explores your existing data sources, schemas, and previous transformations. It reuses what you already have instead of recreating everything from scratch. It learns from your patterns.
This is where the magic is. Express doesn’t just move data – it handles complex, multi-step workflows. It understands dependencies. It suggests next steps. It learns from your existing workflows and applies that knowledge to new requests.
And because it’s built on Nexla’s platform, it inherits:
Express makes developers, analysts, and business users up to 10 times more productive by eliminating the time-consuming process of preparing and integrating data from multiple sources. But it’s more than a productivity tool.
It’s a fundamental shift in who can work with data.
The developer who needs to prototype a new feature can pull the exact data they need without filing tickets. The analyst who spots a trend can combine data sources to investigate without waiting on engineering. The business user who understands their domain can create the data products they need without learning Python.
“Now, anyone who knows the outcome they want can work with data conversationally and reliably, and unlock the data they need, in minutes.”
It’s data engineering transformed into context engineering.
The biggest blocker to AI adoption isn’t the models – it’s the complex, time-consuming process of preparing and integrating data from multiple sources to create context for AI.
Everyone’s talking about AI transformation. But here’s the uncomfortable truth: most companies can’t move fast enough. Not because they lack AI talent or compute resources, but because they can’t get their data ready.
You can’t fine-tune a model without training data. You can’t build RAG systems without context. You can’t deploy AI agents without giving them access to your data ecosystem.
The teams that move fastest with AI will be the ones who can unlock their data fastest.
That’s what Express enables. Not just faster data engineering – faster AI adoption.
We didn’t build Express in a vacuum. It’s built on a platform processing over 1 trillion records monthly for companies that can’t afford downtime or security gaps.
Express is:
You can try it today at express.dev
This is more than a product launch. It’s the culmination of a journey that started with a simple belief: data users shouldn’t have to be data engineers.
We built Nexla to handle data variety. We accumulated knowledge about data systems at scale. We watched conversational AI transform software engineering. And we realized we could do the same for data engineering.
Express is what happens when you combine years of enterprise data experience with the conversational interface that AI finally makes possible.
It’s the bridge between “I need this data” and “here it is, ready to use.”
And I can’t wait to see what you build with it.
Ready to transform how your team works with data? Start building with Express.dev today.
Read the launch coverage:
About the Author: Saket Saurabh is the CEO and Co-Founder of Nexla. He previously worked at NVIDIA, where he learned the art of building elegant solutions to complex problems, creating new applications of GPUs as an entrepreneur inside the company. He later founded one of the first mobile ad platforms, where the need to handle massive real-time data streams pulled him into building large-scale data systems. Saket is passionate about making data accessible to every data user and application.
In this episode of Software Engineering Radio, listen how the Model Context Protocol is transforming AI system architecture, and empowering developers with agentic workflows
