Tacos de Python

• 

  Financials

• 

Inspiration

What it does

• ✅ Heatmaps: Detect areas of interest and where
• ✅ Person Tracking: Track the persons that pass in the areas of your interest, detect where they go, and the areas of interest of them.
• ✅ Personal Media Explorer: Upload your audios and videos to our platform and let us process them for you.
• ✅ Call Insights: Detect if your clients are having a good experience, the sentiments of your calls, and get the information that matters about them.
• ✅ Interaction Analysis: Check out how you interact with other persons and let us do the analysis of the interaction for you.

How we built it

Challenges we ran into

We had some problems using Open Vino, as it was only available through Intel GPUs, as we only had one computer that had an ARC GPU. Once installed we ran into more library problems, which we had to modify some internal utils until they all worked. Another struggle we had was to effectively use FRIDA, which was not easily accesible through our college's WIFI.

Accomplishments that we're proud of

1. 📊 Retention Analysis Client Hero offers a detailed analysis system on customer retention time in different areas of the company, allowing the identification of areas for improvement and critical points in the service.

2. 📅 Interactive Calendar With our interactive calendar, administrators can select any day to analyze customer service performance in that specific period, earning valuable insights.

3. 🕒 Hourly Retention Visualize how customer retention changes throughout the day. The bar graph shows the maximum retention times per hour, from 9:00 AM to 6:00 PM. This helps identify peak hours and improve resource management.

4. 🗾 Management of Service Zones Client Hero allows data to be segmented by focus area, providing custom insights for each area, making it easy to optimize services where necessary.

Startup's Financials

We explored this project as a viable solution that could quickly find product market fit, as it includes multiple services that empower not only SMEs, but also big corporations, as the unique offering includes features that have actually been validated with some companies. We decided to create a financial projection for the startup, focusing only on SMEs, as this could ensure a positive turnover in the first year. The other path is to offer more services at a loss and finance it through some venture investing. This path is optimal to quickly gather market.

What we learned

We learned a lot about how OpenVino works, the arquitecture and how to use model convertors, which at the end optimized the performance running in Intel's hardware. We also learned how to use FRIDA and the possibilities such LLM provides.

What's next for Tacos de Python

Built With

• chroma-db
• computer-vision
• fastapi
• firebase
• frida
• langchain
• mongodb
• openai
• openvino
• python
• rag
• streamlit
• ultralytics