Stonk Hub

A finance-focused web application presenting an AI-assistant workflow for historical market data, with account entry points, plan comparisons, and a clear product onboarding experience.

Making a historical market-data product understandable through a conversational SaaS experience.

Stonk Hub presents a conversational approach to historical market data. The public experience introduces an AI-powered assistant as the entry point, then explains the service through a pricing comparison, credit allowances, data coverage, and a detailed FAQ.

The application is associated with the chatbot-project repository. Its frontend uses React, TypeScript, Vite, Tailwind CSS, and shadcn/ui.

Market-data products expose unfamiliar distinctions: asset classes, historical depth, intraday access, exchanges, and usage credits. If those details are hidden behind the chat interface, a user cannot make an informed decision about whether the product fits their task.

The design challenge was to make the conversational idea approachable while keeping the underlying product boundaries visible.

1. Review assistant.
2. Compare data access.
3. Choose account entry.
4. Request market data.

The flow above describes the product journey. The following sections explain the implementation and the boundaries of this case study.

The public product surface brings onboarding and commercial explanation together in a React interface. The repository also contains Supabase resources, while the public deployment maps to the finance-focused Stonk Hub identity.

A useful product boundary is the distinction between explaining the assistant and executing a data request. Pricing and FAQ content set expectations before the user reaches the application. The frontend stack supports a consistent set of components for these views.

The screenshot review did not exercise the model backend or purchase a plan, so it does not establish which provider serves the data or how reliably the assistant fulfills every request.

Technology stack

A conversational interface can hide complexity, but it cannot remove the need to define available data and usage limits. Those constraints should stay understandable in the product.

The public website and repository stack were inspected. Live market-data fulfillment, AI response quality, subscription billing, and authenticated-user permissions were not transactionally verified in this portfolio review.

AI product design benefits from clear expectations. The assistant is the interaction model, while plan boundaries, coverage, and account behavior are part of the product contract the interface must explain.

Open the application