StockSherlok: Your AI Stock Detective

AI tool that uncovers high-growth mid-cap stocks before they trend, with quick insights and easy comparisons.

Comment

Inspiration

StockSherlok is an autonomous AI research agent designed to uncover fast-growing, under-the-radar companies that traditional stock dashboards overlook. While everyone invests in the same 10–20 famous companies, thousands of mid-cap and emerging tech players (AI, robotics, biotech, cybersecurity, hardware) are quietly scaling and often impossible to discover manually.

StockSherlok works like a financial detective. It fetches market data, analyzes patterns, and summarizes growth signals so users can track early opportunities without needing expert knowledge.

What it does

StockSherlok acts like an AI stock detective. You type a question, and it instantly analyzes multiple mid-cap companies, compares their performance, and shows the top picks with charts, trends, and a simple summary. It makes stock research feel easy, visual, and fast.

How we built it

I built a Python backend that calculates real stock metrics, an AGI/OpenAI agent that ranks companies using reasoning, and a React dashboard (designed in Lovable) to show everything in a clean, interactive way. Telnyx Voice and Perplexity outputs helped me create realistic insights. I built every part solo the agent logic, scoring, charts, and UI.

Challenges we ran into

The hardest part was connecting everything smoothly, metrics, charts, the agent, and the UI. Stock data can be inconsistent week-to-week, so I created a static demo mode for stability. Designing a simple but smart-feeling interface also took a lot of iteration. Getting the agent to give clear explanations without overwhelming the user was another challenge.

Accomplishments that we're proud of:

I’m proud that the whole system works end-to-end: type a query → get ranked stocks → see comparisons → read a clear summary. The dashboard feels polished, the outputs look real, and the experience feels like a real AI financial assistant. Building this completely solo in a short time is something I’m really proud of.

What we learned

I learned how to design a vertical agent, how to combine reasoning with real metrics, and how to structure clean backends and frontends that talk smoothly to each other. I also learned a lot about simplifying complex data so anyone can understand it. And I gained experience balancing “real mode” and “demo mode” for a smooth judging experience.

What's next for StockSherlok: Your AI Stock Detective:

I want to add more financial signals, analyst sentiment, and price alerts. I also want to support voice conversations, build a portfolio simulator, and add support for trending sectors like AI chips, cybersecurity, and robotics. Long-term, StockSherlok could become a personalized AI research assistant for everyday investors.

Built With

Share this project:

Updates