L'investique

• 

  pick a ticker

• 

  get a chart

• 

  and analysis of it (summary, news and tips)

• 

  bonus: backtesting with performance %

Inspiration

Tens of millions of traders globally spend hours manually analyzing stock charts, which is time-consuming and error-prone. Recognizing this challenge, we saw an opportunity to revolutionize chart analysis by leveraging multimodal reasoning in an agentic pipeline, enabling automation that saves time and minimizes human bias in trading decisions. Our personal experience played key role. Nur specifically spent over 7 years in trading and having worked at London Stock Exchange and $100bn investment fund - he has faced the issue again and again.

What it does

L'Investique automates stock chart analysis for traders. By entering a stock ticker, the tool pulls stock charts, analyses key trading indicators such as Bollinger Bands, MA, RSI etc using Multimodal models, and provides a concise summary of key insights, trends, and signals with some trading tips to the user. As a result, it removes the need for manual interpretation, reducing human error and cognitive load.

How we built it

We used a combination of APIs (tradingview & bravesearch) to pull stock data and generate technical charts. Then, leveraging the Mistral's Pixtral 12B we analysed the chart images and extracted key data. From there, we built an AI-driven summarization tool to convert this data into easy-to-understand insights for the trader. The system was developed & run using Python.

Challenges we ran into

The main technical challenge we faced was that our decision-making process relied solely on a single financial chart, disconnected from external factors, which limited our ability to understand the underlying reasons behind market movements. We addressed this by integrating a news API, allowing the agent to access news feeds relevant to the chart’s time period.

Accomplishments that we're proud of

We automated a manual process that traditionally takes hours and requires expertise, creating a tool that is useful for both beginner and expert traders. More to that, there is strong evidence that novice traders oversimplify trading as they dont want to manage complex charts (we have solved that now and brought more reasoning to decisions). Positive feedback from traders network was very encouraging.

What we learned

We learned how powerful Multimodal models can be when applied to financial data, and the potential of automation in trading. We also discovered the importance of delivering insights in a way that’s easy for users to understand.

What's next for L'investique

We plan to add more customization options, real-time alerts, and integrate with trading platforms (and user's trading accounts). Scaling to other markets like crypto (already tested), fx and commodities is now ahievable.

Link to our deck

Built With

• bravesearchapi
• mistral
• pixtral12b
• python
• tradingviewapi