An AI copilot that is not allowed to guess
Adding language-model features to a Power BI-style dashboard builder, where every column the model names is checked against the data model before anything runs.
Farakav · Golrang System · 2025–now
Nine years of React and TypeScript on enterprise products in process mining, business process management, digital banking and eLearning, and lately the AI features and .NET services behind them.
Open to remote roles and to relocation in Europe. One month’s notice.
Five problems I keep meeting, each with its case study:

Adding language-model features to a Power BI-style dashboard builder, where every column the model names is checked against the data model before anything runs.
Farakav · Golrang System · 2025–now

The digital Open Account journey for a consumer bank, where every extra step loses a share of the people who started.

Adding custom elements and behaviour to a mature open-source workflow modeller through its own extension points, without forking it.

Finding the endpoint behind an analytics platform’s slowest workflows, and moving hundreds of per-field queries to the moment someone opens an item.

Turning a German self-study language platform into a learning management system, and answering one question three times: who logs in, and what do they see?
B.Sc. Information Technology, Islamic Azad University, South Tehran Branch, 2013 – 2019.
The full CV, with every role in detail.
Farakav: the process-mining platform, and a rewrite of its dashboard builder with AI features and a .NET back end.
Business process management suite: BPMN-IO and FormJS extensions, an access-control platform.
Bajet, a consumer banking PWA: the Open Account journey, onboarding 600+ verified users a day.
A language-learning platform, rebuilt in 2025 as a learning management system.
React and WordPress builds; performance, SEO and accessibility work.
Startup products in AngularJS, in a cross-functional Agile team.
A plausible column name that does not exist is worse than an error, because people trust the chart that comes out of it.

Most of my nine years have gone into enterprise products where the hard part is what happens to an interface when real data volume arrives: a process-mining platform rendering analytical datasets, a business process suite whose workflow editor had to do things its library was never designed for, and a consumer bank whose onboarding flow had to work for six hundred people a day.
In all of them the frontend is the product. There is no fallback interface or power-user CLI, so if a table locks up while it sorts, the feature is broken.
I work in React and TypeScript, long enough to have opinions about state management that I hold loosely. I have worked remotely with German clients since 2019, which has made me better at writing things down.
At Golrang I now build AI features as well as interfaces. In the Farakav dashboard builder, language models propose charts and DAX measures, explain and repair expressions, and answer questions about a dashboard by running real queries. Each feature checks what the model returns against the data model before anything runs. I also use AI tools in my own work, and I review what they produce the way I would review a colleague’s pull request.
That is also where I want to keep growing. Alongside the frontend, I am learning the .NET and database side of the products I work on, and the AI features built on top of them, and I am looking for a team where I can do more of both.