Stem4All

• 

  Cover page

• 

  Architecture Diagram

• 

  demo

• 

  demo2

• 

  inspire more women to join stem

• 

  Demo 5

• 

  Data Engineering on snowflake

• 

  demo 6 : give me the end to end migration strategy for mysql to snowflake migration

• 

  ease of using AI on snowflake

Inspiration

Women remain significantly underrepresented in the fields of data engineering and artificial intelligence (AI). As of 2020, women held approximately 26% of data and AI-centered jobs, with a notable gender gap at the executive level. FACEBOOK

In the United States, women account for less than 23% of the tech workforce, encompassing roles in data engineering and AI. This underrepresentation is attributed to factors such as unconscious bias, lack of mentorship, and limited access to resources. REUTERS

Globally, women comprise only 22% of AI talent, with even lower representation at senior levels, occupying less than 14% of senior executive positions. INTERFACE

In the United Kingdom, women make up 22% of AI and data professionals and 18% of users across the largest online global data science platforms. THE ALAN TURING INSTITUTE

These statistics highlight the persistent gender disparities in data engineering and AI, underscoring the need for targeted initiatives to promote inclusivity and diversity within these fields

What it does

At STEM4All, our mission is to foster inclusiveness, diversity, and innovation in the STEM fields, empowering individuals—especially women—to become changemakers in their communities and industries. We believe in breaking barriers and ensuring that everyone, regardless of gender or background, has the opportunity to excel in Data Engineering and Artificial intelligence.

This application is a Streamlit-based Question-Answer System leveraging Retrieval-Augmented Generation (RAG) with Snowflake as its core infrastructure. It integrates Cortex Search for efficient information retrieval and the Mistral Large Language Model (LLM) for generating accurate, context-aware responses to user queries

How we built it

User Interaction:

The user accesses the front-end interface built on Streamlit. The user submits a query via the Streamlit app. Search and Retrieval (Cortex Search):

The query is passed to Cortex Search, which performs an intelligent search on the indexed documents to find the most relevant information. Cortex Search returns the relevant documents or data based on the user's query. Generation (Mistral LLM):

The retrieved documents are passed to the Mistral LLM (mistral-large2) running on Snowflake Cortex. The model processes the documents and generates a detailed response to the user’s query, using the retrieved context. Displaying the Results:

The generated response is sent back to the Streamlit app, where it is displayed to the user in a user-friendly format.

Challenges we ran into

collecting the right documents. getting the right prompts

Accomplishments that we're proud of

STEM4All fosters inclusiveness, diversity, and innovation in the STEM fields. It is a platform that helps women get access to Data Engineering and AI from the comfort of their homes and encourages them to join stem jobs.

What we learned

Stream Lit , Rag, True Lens, Mistral , snowflake

What's next for Stem4All

Adding extensive interview preparation and suggesting learning path based on the users skill level. Adding more documents on Gen AI and system design Add features where the user can share their interest in a certain area and skill level and Stem4All with suggest them a customized learning path.

Built With

• mistral
• rag
• stream-lit
• true-lens