Viserra: Simplify stock research for retail users with AI-powered data

Aug 2025 - Sep 2025

A stock research platform built in Lovable that helps users explore stocks based on their interests and risk awareness.

My Role

My Role

My Role

Product Designer, Developer, User Researcher

Worked With

Worked With

Worked With

Startup Founder

Timeline

Timeline

Timeline

Aug 2025 - Sep 2025 (1 Week)

Aug 2025 - Sep 2025 (1 Week)

Skills & Tools

Skills & Tools

Skills & Tools

Rapid Prototyping, Vibe Coding, Lovable, Claude, Gemini

Summary

This is a working build that leverages AI and LLMs to screen stocks, explain complex financial metrics in plain language, and personalize insights based on user preferences. The system demonstrates how AI can turn overwhelming research into accessible insights.

Overview

Overview

Overview

A prototype to quickly test the market.

It goes beyond static screens on Figma — it’s a functional build:)

Viserra is an early-stage fintech startup focused on real estate investing (REITs). The platform brings proprietary financial data into one place, so investors don’t need to search across multiple reports or calculate metrics on their own.

This project is a prototype for an upcoming feature that helps users screen stocks based on their goals, risk tolerance, and industry preferences…etc.

A stock research platform that cuts down hours of research with AI, screening stock candidates and explaining complex financial metrics in plain language.

Workflow

Workflow

Workflow

💻 Frontend

Interactive UI

Interactive UI

🗄 Backend

Database & Hosting

Database & Hosting

📊 Financial Data

Stock Price, Company Info

Alpha Vantage/FMP API

Stock Price, Company Info

Alpha Vantage/FMP API

Stock Price,

Company Info

Alpha Vantage/FMP API

🧠 AI + LLM

Personalized Analysis

Personalized Analysis

Key Components

Key Components

Key Components

  1. Onboarding Quiz

Questions about goals, risk comfort, industry preferences, and investing habits help identify which companies align with users’ expectations.

  1. Tailored Candidates

From a user-centric perspective, the system screens stock candidates, integrates API data, and leverages LLMs to generate a brief personalized analysis.

Key industry metrics are explained in plain language, helping users avoid complex financial jargon.

  1. Explore Details

Users can click View Details to access a deeper layer of analysis, including extended explain of financial metrics, historical trends, and AI-generated insights.

The About the Company section provides background to help users understand company's business.

  1. Add to Watchlist

Users can save stocks they are interested in to their Watchlist, allowing them to revisit these candidates later with updated data.

What are we solving?

What are we solving?

Problems & User Pain Points

Problems & User Pain Points

Problems & User Pain Points

Many retail investors, especially beginners, feel overwhelmed by the amount of research and financial metrics required before making investment decisions. Even after spending time on research, many still don’t know what is suitable for them.

Time Consuming

"Do your own research"—but digging through reports, news, and articles can take hours.

We identified the need to distinguish between external disruptions (e.g., notifications) and internal, habit-driven interruptions to avoid attention mechanisms that disrupt focus.

Knowledge Gap

Hard-to-understand metrics leave beginners overwhelmed and unsure which ones are truly important to focus on.

Limits of Existing Advisors

Financial advisors can offer personalized strategies but charge fees, while robo-advisors often push their own products and lack objectivity.

Opportunity

Opportunity

Opportunity

Help investors save time and reduce confusion by turning complex financial data into personalized insights that support their investment research process.

Click the image and try it out!

Reflection

Reflection

Reflection

What I Learned from AI Prototyping…

  1. Empowering Multi-Role Work

Lovable let me act as a PM, designer, and engineer all at once. From a designer perspecitve, vibe coding tools like this can accelerate ideation, save us significant time on wireframing and prototyping, and even make building an MVP possible within days.

  1. The Cost of Speed…

AI often misinterpreted prompts or fabricated results, and I only realized it after looking deeper. For example, it invented data when an API failed. Since vibe coding is still mostly one-directional—where users give prompts but the AI doesn’t ask clarifying questions or highlight limitations like a real engineer would—it often tries to “solve” problems in unintended ways. So I still spent days debugging. So far, I believe human judgment is still necessary to do QA.

Reflection

Reflection

Reflection

What I Learned from building a financial product…

  1. Empowering Multi-Role Work

Lovable let me act as a PM, designer, and engineer all at once. From a designer perspecitve, vibe coding tools like this can accelerate ideation, save us significant time on wireframing and prototyping, and even make building an MVP possible within days.

  1. The Cost of Speed…

AI often misinterpreted prompts or fabricated results, and I only realized it after looking deeper. For example, it invented data when an API failed. Since vibe coding is still mostly one-directional—where users give prompts but the AI doesn’t ask clarifying questions or highlight limitations like a real engineer would—it often tries to “solve” problems in unintended ways. So I still spent days debugging. So far, I believe human judgment is still necessary to do QA.

Financial Regulations & Compliance

A challenging part of building this is balancing valuable insights with the necessary legal disclaimers. Since this prototype focused on demonstrating the user flow, disclaimers such as ‘All stock data and financial metrics are for educational and demonstration purposes only. Past performance does not guarantee future results.’ were included.

Financial Regulations & Compliance

A challenging part of building this is balancing valuable insights with the necessary legal disclaimers. Since this prototype focused on demonstrating the user flow, disclaimers such as ‘All stock data and financial metrics are for educational and demonstration purposes only. Past performance does not guarantee future results.’ were included.