FMartull / oxford-analytics-session

This repository contains code files from the Oxford University AI Course AI session from January 2022

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Oxford University Analytics Session

Saturday 8th January 2022

Storyline:

Tailwind Traders are DIY, Home and Garden store in and around the London area. The COVID19 pandemic has really shook up the business and priorities, they need deeper insight into where to focus precious time, money and resources in investing in their future. Tailwind traders have employed us as their data science consultant team to help them form a data driven plan for the next financial year

Session Detail:

The data science team have been tasked with:

  • Understanding the impact of COVID 19 and change in peoples lifestyle patterns in London Boroughs
  • Recommend the top 3 areas for Tailwind Traders stores to consider for future investment
  • Share cutting edge Internet of Things solutions for understanding footfall in and around the business
  • Create AI solutions that support operational efficiency and future predictions into the state of the business

Session Breakdown:

Go to section 1

  • Section 2: Now we understand the data, create and share Power BI reports/dashboards with business users at Tailwind Traders. Tell a story back to the business of the data and recommend some next steps for further investigation

Go to section 2

  • Section 3: We are now ready to introduce how to employ various IoT technologies in and around our stores to analyze footfall using a production-grade architecture that allows for remote management, over the air updates, and customized reporting. In this section, Paul DeCarlo will share insights on implementation patterns for Computer Vision using Azure IoT Edge, Azure IoT services, and NVIDIA Jetson Hardware.

Go to section 3

  • Section 4: Further recommendations will include forecasting and operational efficiency which can be achieved using AI solutions
    • Starting with an Azure Machine Learning code-based solution to show future forecasting of activity in London boroughs which can help Tailwind Traders make decisions on where to place new stores given the pandemic changes to activity
    • Talking about the pro's and con's of using a code-based solution and introducing the no-code opportunity with AI Builder as a part of the Power Platform
    • Building a forms recognizer solution for analysing Tailwind Traders receipts and invoices with no-code (using AI Builder and Power Automate)

Go to section 4

Case Study and Student Exercise:

(same storyline continued…)

Tailwind Traders are DIY, Home and Garden store in and around the London area. The COVID19 pandemic has really shook up the business and priorities, they need deeper insight into where to focus precious time, money and resources in investing in their future. Tailwind traders have employed us as their data science consultant team to help them form a data driven plan for the next financial year.

After the initial feedback sessions sharing the Google Activity data by London Borough, possible IoT architecture solutions and AI solutions that support operational efficiency and future predictions - the Tailwind Traders team have funded another round of consultancy and investigation work. This time they want to focus on marketing efforts - how do they gain net new customers.

One option, as they are based primarily in London, is they want to get in front of new customers whilst they are travelling about their busy days and take a break by stepping into a store or browsing online.

As the new Data Science lead on the project you are tasked with:

  • Explore the Public Transport Journeys by Type of Transport - London Datastore open dataset to understand types of journeys people are making around London using Power BI.
  • Add a page to the Power BI report shared with Tailwind Traders in round one of investigations
  • Recreate the forms recognizer solution for analysing Tailwind Traders receipts and invoices with no-code using AI Builder and build this functionality into a Power App
  • BONUS: explore https://api.tfl.gov.uk to see if you can connect any further transport and activity patterns around London for Tailwind Traders

Go to Student Exercise

About

This repository contains code files from the Oxford University AI Course AI session from January 2022

License:MIT License