Your team already turns raw data into pipelines, datasets, and dashboards inside OpenHEXA. The hard part has always been the last mile: packaging that works into something other people can actually open, click through, and use. Building a data product often meant handing off to engineers, wiring up access controls, and waiting for a deployment.
That last mile just disappeared. OpenHEXA now ships a native AI assistant that builds web applications from a plain-language description, using the data already sitting in your workspace. Describe what you want. Watch it get built. Publish when you’re ready.
Webapps are interactive applications hosted directly inside your OpenHEXA workspace. They come in three flavors:
Each one lives alongside your pipelines and datasets, governed by the same workspace it belongs to. No separate hosting, no separate login, no separate stack to maintain.
Here’s where it gets fun. The assistant sits right next to the code editor and does the building. Tell it what you need in ordinary language, and it draws on the datasets, files, and pipeline outputs already available in your workspace to produce a working app.
Want a map of case counts by district, refreshed from last night’s pipeline run? A survey-completion tracker your on-site teams can check on their phones? A one-page report that pulls the latest figures every time someone opens it? Describe it. The assistant writes the code, connects the data, and hands you something real to react to. Iterate by talking to it. Refine by editing directly. You stay in control of the result, and the finished app is fully owned by your team.
This shortens the distance between an idea and a shipped tool. The person who understands the data no longer waits on someone else to expose it.
Speed means nothing if it puts your data at risk. This assistant was designed for organizations that hold themselves to a high bar, so control is built into every step.
Nothing goes live before review. Every change is versioned. Apps publish only when your team decides they’re ready, so work in progress stays in progress until you say otherwise.
Each app sees exactly what it should. Every app receives scoped API access to workspace assets. It shows its users precisely the data they’re cleared to see, and nothing more. Permissions follow the workspace, not a copy of the data floating somewhere it shouldn’t be.
Your team owns the output. The assistant builds the app, but the code, the data, and the finished product belong to you. Inspect it, extend it, hand it to your engineers. No lock-in, no black box.
The native assistant is one way in. It isn’t the only one. OpenHEXA also ships an MCP (Model Context Protocol) server, so the AI tools you already use, like Claude and ChatGPT, can build webapps straight from your workspace. Point your assistant at it and ask for a dashboard: it creates and edits the static webapp, wired to your datasets and pipeline outputs, then you publish through the same governed workflow. Same result, from whatever assistant you prefer.
See it for yourself. Start a free 30-day trial, spin up a workspace, and describe the first app you have in mind.
Every AI feature is included in the trial, at no extra cost. The native assistant, the MCP server, the full webapps workflow: all of it is on from day one, so you can test the exact experience your team would ship with. No add-on, no upgrade, no waiting.
Exposing your work to the world has never been easier. Go build something.
