Microsoft Fabric Updates Blog

Announcing the winners of “HackTogether: The Microsoft Fabric Global AI Hack”

We’re excited to announce the winners of HackTogether: The Microsoft Fabric Global AI Hack!

What is #HackTogether?

In this HackTogether, we had the opportunity to build innovative AI solutions with Microsoft Fabric. The event challenged over 640 registrants with prize awards for top submissions (including one grand prize that includes passes to attend the Microsoft Fabric Community Conference)! Over the course of the two-week event that ran from February 15 to March 4, we delivered special livestream sessions that included a kick-off event with Kim Manis (Vice president of Product, Microsoft Fabric & Power BI) and taught hackers how to build custom object detection models, RAG applications with Azure Open AI, and create AI prediction models using Microsoft Fabric. To ensure that hackers had access to the latest and greatest AI services from Microsoft, we provided an Azure Open AI proxy service that allowed hackers to incorporate Azure Open AI without the need for an active Azure subscription. Over the course of the hackathon, hackers made over 54,000 requests to the Azure Open AI proxy!

The winners

We received 50 Hackathon project submissions from over 100 registrants, participating from every corner of the world. Our judges undertook the extremely difficult task of selecting four prize-winning submissions aligned to Hackathon categories including: Best Real-World AI application built with Microsoft Fabric, Best use of Microsoft Fabric + Azure OpenAI, Best use of Copilot for Microsoft Fabric, and an additional Grand prize award for the best of the best!

Our judging process considered overall alignment to the corresponding Hackathon category, degree of innovation / impact to the real-world, clarity of documentation and reproducibility, and overall quality of a hacker-provided video demonstration for all submissions.

Our judges were blown away by the breadth, depth, and overall quality of submissions. Each and every project had its own unique elements that demonstrated Microsoft Fabric is truly the Data and Analytics platform for the era of Artificial Intelligence. As a result, we will also be sharing honorable mentions. Even though there can be only one overall prize winner for each category, we feel it is important to share with you how much dedication went into selecting the winners and give our standout submissions the spotlight they deserve.

🏆 Grand prize winner: PIInovators

PIInovators is a cloud-native data solution developed in Microsoft Fabric, integrating OpenAI for document analysis, particularly for detecting Personal Identification Information (PII) in files and images. The solution classifies documents into compliant and non-compliant categories, further categorizing them based on predefined types (Delivery, Personnel, Online, Continual, Communication). The solution is based on a medallion architecture where data is stored in three zones (Bronze, Silver, Gold) within Microsoft Fabric Lakehouse. Subsequently, the data is prepared for analytical use in Power BI reports.

Project Repository URL

https://github.com/esddata/piinovators.git

Project video

🏅 Best Real-World AI application built with Microsoft Fabric: The Foodwaste Hack

In this hack we want to show you how you can combine Powerapps, Fabric and AI to minimise food waste in your kitchen.

Ever looked into your fridge and thought “darn, should have cooked that chicken sooner?” Worry no more. we have a hack for you.

In our github repo you will find everything you need to create a power app to help track food you have bought, assisted by your own lakehouse with millions of product descriptions, use AI to help with setting expiration dates, and again use AI to recommend some yummy recipe ideas on what to do with your ingredients. The best part? A reflex trigger in Fabric will know when your items are about to go off and automatically trigger your own master chef to suggest some great recipe ideas and email them to you before it is too late!

Project Repository URL

https://github.com/AllgeierSchweiz/aihackers

Project video

🏅 Best use of Microsoft Fabric + Azure OpenAI: Building an Analytics Solution over a Text Document Repository

This project leverages the power of Microsoft Fabric and Azure OpenAI to enrich and analyze a repository of text-based data. So many companies have valuable unstructured data that can now be unlocked with Azure OpenAI. In this project, Azure OpenAI enriches the data by performing entity extraction, text summarization, text classification, and embedding generation for semantic similarity. The end result is an analytics solution with Notebooks and Power BI over the document repository that can be used by both developers and business users!

Project Repository URL

https://github.com/BrightonKahrs/project-gutenberg-analysis/tree/main

Project video

🏅 Best use of Copilot for Microsoft Fabric: Accounts Payable Payment Prediction

Our project uses Accounts Payable data from an ERP system to predict the payment status of unpaid invoices – specifically whether they will be paid early, late, or on-time. This data in moved through a “medallion” lakehouse architecture and modeled in various notebooks. The results of the predictive model are presented alongside traditional AP reporting in Power BI. Additionally, we generated our calendar table and semantic model descriptions using Copilot. This project is designed to fulfill a real-world need and will be further developed into a production-ready solution for our organization. Thanks for considering our submission!

Project Repository URL

https://github.com/aboerger/Fabric-Hackathon-AP-Payment-Prediction

Project video

💝Honorable Mentions

We received many high-quality submissions and encourage you to take a look at the hackathon entries to see if there is something relevant to your line of work or interests. Each of these submissions has provided value not just for the hackers but for anyone interested in looking into how their particular problems were solved using Microsoft Fabric workloads along with AI methodologies. We share 12 of our top picks below for your perusal.

🪙Fabric SQL DB audit

The solution proposed involves the collection of SQL audit logs and their near real-time streaming in a simple way, into Microsoft Fabric thanks to its simplicity

This efficient process ensures that logs are stored systematically and securely.
Furthermore, the solution offers additional capabilities that are critical for data analysis and management.

Thanks to Fabric’s capabilities, include the ability to analyze audit events with LLM and the compatibility with various data scientist tools.

This solution, therefore, not only streamlines the data collection process but also enhances data utility, making it a valuable addition to any data management strategy.

Project Repository URL

https://github.com/gianlucadardia/FabricHack.git

Project video

https://github.com/gianlucadardia/FabricHack/blob/main/video/FabricHack.mp4

🪙Anomaly Detection in Email Datasets

We have crafted a cutting-edge anomaly detection solution tailored for email datasets, exclusively leveraging MS Fabric. Emails, being vast and unstructured data, pose a formidable challenge for processing and comprehension. Harnessing the power of Azure OpenAI, we have ingeniously devised a solution that transforms this unstructured data into a structured format, ideal for generating insightful reports and visually engaging representations such as Power BI reports. Additionally, through adept utilization of MS Fabric, our solution seamlessly triggers pipeline processes at regular intervals to efficiently handle new subsets of emails.

Project Repository URL

https://github.com/aokic1/ms_fabric_hackathon_2024

Project video

🪙Fueling Innovation and efficiency using Microsoft Fabric

This proof-of-concept has been developed for global tank manufacturing and inspection company. The solution aims to revolutionize the manufacturing, repair & maintenance(R&M) and inspection process by leveraging Microsoft Fabric.

The state-of-the-art cloud data platform is intricately designed to seamlessly ingest data, process and empower advanced analytics thereby driving operational excellence.

The data platform offers a comprehensive view across ingestion, processing, and consumption stages, while also addressing data governance aspects such as Data Security, Data Quality, and Data Lineage etc.

Project Repository URL

https://github.com/darshanpv/MS_Fabric_Cloud_Data_Platform

Project video

🪙Ask-MSlearn

Our idea for this competition was to provide an interactive chat-interface to the articles on mslearn. The mslearn database is a vast resource with information on many different Microsoft products and tools. It is navigable through tags and collections like modules and learning paths, but it can still be tricky to find just the article you need in a particular situation since they overlap on certain topics.

We wanted to make it easier to filter through the modules in the mslearn library, and we wanted to extend the functionality from just finding and reading up on certain topics, to being able to automatically find all modules related to some topic, so that it’s easier to compare the different tools, products and methods. We do this by creating an interactive chat-like experience where you explain your problem/situation to an AI and get recommendations based on the knowledge in mslearn.

For example you could:

  • Post your resume to get suggestions of tools and products that you might be interested in.
  • Describe a workflow and get suggestions for tools and products that might help you get the work done.

Project Repository URL

https://github.com/RikardMartin/mslearn-smart-chat

Project video

https://github.com/RikardMartin/mslearn-smart-chat/blob/main/docs/Ask-MSlearn-video.mp4

🪙Daily NYC Vehicle Collision Reporting

In New York City in the United States, a standard police report is required when a vehicle collision occurs where someone is injured, killed, or if there is at least $1,000 worth of damage. This data is collected and published online as an open data source at NYC Open Data and is updated daily.

The scenario is to provide Power BI reports that are updated daily to provide descriptive analytics and insight into location, contributing factors, and vehicle types for vehicle collisions using Fabric and CoPilot. This solution also provides foundations for other personas like Data Analysts and Data Scientists for future work.

Project Repository URL

https://github.com/cameron-thorne/mshackathon_nyc_collision/tree/main

Project video

https://github.com/cameron-thorne/mshackathon_nyc_collision/blob/main/Walkthrough/3_3_2024%2C%209_22_16%20PM%20-%20Screen%20-%20Walkthrough.webm

🪙 Parking management using Microsoft Fabric and edge workloads

We choose the parking management scenario where gates equipped with ANPR enabled cameras could use AI to identify and stream cars coming in and out. These gates can operate locally on their own, feeding panels indicating how many spaces are estimated to be available, while streaming data in real-time to Fabric for further aggregation, processing and analysis. IoT Operations (installed on an Azure Arc enabled Kubernetes cluster) enable a two-way communication between devices that may use old protocols and would be hard to upgrade in a secure and very easy to set up way.

Project Repository URL

https://github.com/danuw/fabric-aihack

Project video

Automating synthetic data creation & reporting using Microsoft Fabric

When considering privacy protection in the context of an outsourcing company, synthetic data generation becomes particularly relevant due to the sensitive nature of the data involved. Outsourcing companies often handle data from clients that may contain personally identifiable information (PII), financial records, proprietary business information, and other sensitive data.

By leveraging synthetic data generation techniques, outsourcing companies can effectively manage and mitigate the privacy risks associated with handling sensitive data, thereby building trust with clients, enhancing data security, and ensuring regulatory compliance.

Project Repository URL

https://github.com/corneliascode/Automating-synthetic-data-creation-reporting-using-Microsoft-Fabric-

Project video

🪙Economic Impact Analysis using Copilot

We will use Microsoft Fabric to perform Economic Impact Analysis

  1. We will analyze various Factors of the Economy like unemployment, GDP, cpi, mortgage (copilot for data engineering)
  2. We will predict the mortgage rates for the next 3 quarters. (copilot for data engineering)
  3. We will analyze sales of 10 companies during this period. (co pilot for data engineering )
  4. Build Power BI Reports (copilot for Power BI)
  5. Send data from LakeHouse to SQL Server using Data Factory (For Requirements where other downstream systems need data in their env) (using copilot for DataFactory)

Project Repository URL

https://github.com/shreyasrastogi/FabricAIHackathonSubmission

Project video

🪙CO2SNAP

CO2SNAP is an innovative application designed to assess and provide insights into the carbon footprint of grocery shopping. It utilizes a robust architecture integrating various technologies and services to capture, store, process, and analyze images of shopping receipts to estimate the CO2 emissions associated with purchased items.

Project Repository URL

https://github.com/robinhubner/microsoft-fabric-gloabl-ai-hack

Project video

🪙MS Fabric Flood Predictions

For many people across the world, risk of flooding is a yearly concern, impacting both home life and work life when roads become blocked off and streets are flooded. Floods are a growing problem, with climate change and a reduction in green space increasing the risk of extreme weather events and water runoff.

Our idea for this hackathon was to use the range of functionality within Fabric to surface river level and rainfall data to allow a holistic view of potential future flood risk within a given area. This could then incorporate other data sources such as news reports or Twitter/X posts.

We aimed to build a flood prediction system using input from the UK Environment Agency (EA) public APIs. The dataset contains:

  • Flood Warnings / Events
  • Flood Area mapping data
  • Stations for measuring river height and flow levels, rainfall and temperature.

Project Repository URL

https://github.com/methodsanalytics/ma_fabric_hack_together_2024/

Project video

https://github.com/methodsanalytics/ma_fabric_hack_together_2024/blob/main/VideoSubmission/Fabric%20Hackathon%20Video.zip

🪙Alert-O-Monitor

“Alert-O-Monitor” is a one-stop solution for managing end-to-end flow of Early Warning Signals in Banks to prevent slippages of accounts into NPA (Non-Performing Asset) category. It involves “Alert Generation”, “Alert Monitoring”, “Alert Resolution” as well as “Alert Prediction” giving a complete 360-degree comprehensive solution.

Business Outcomes :

  1. Reduction in “Provisioning for Credit Losses” leading to increase in Profitability of Banks
  2. Decrease in Impairment of Assets leading to “Better Asset Quality”
  3. Decrease in instances of Frauds in Banks
  4. Improved Credit Decision-Making and Underwriting process
  5. Faster TAT (Turn Around Time) for Loan Recovery
  6. Accountability and Transparency within Bank to ensure timely identification and closure of Alerts

Project Repository URL

https://github.com/shaleen410/Alert-O-Monitor

Project video

🪙San Francisco-Marin Food Bank Household Participation Model

 Using inferential modeling to uncover which factors make a participant more or less likely to attend a food distribution. These drivers of attendance can then be communicated to business owners and recommendations for updating programming can be made.

A logistic regression model is created with various participant attributes and the coefficients from this model are then written to a delta table which is loaded into Power BI. Using Power BI users are able to view the odds and associated standard errors of the different categorical and numerical predictors in relation to likelihood of attendance.

Project Repository URL

https://github.com/Data-SFMFB/SFMFB-Fabric-Hackathon-Submission

Project video

https://github.com/Data-SFMFB/SFMFB-Fabric-Hackathon-Submission/blob/main/HackathonRecording.mp4

🪙Eyes on Future

“Eyes on Future” is an project under the Microsoft Fabric Global AI Hack, focusing on environmental sustainability through the lens of data analytics and artificial intelligence. It aims to deliver actionable insights into meteorological conditions by analyzing current and historical weather data to forecast future climate trends and assess the impacts of ecological initiatives.

Objectives:

  • To analyze current weather conditions and compare them with historical data.
  • To forecast future meteorological trends and assess the impact of environmental initiatives.
  • To leverage the full suite of Microsoft Fabric’s analytics and AI capabilities for comprehensive environmental analysis.

Project Repository URL

https://github.com/anwarmelh/Eyes-On-Future

Project video

🪙Fabric for Canadian Permanent Residence

This project designs an end to end solution for Canadian Permanent Resident Applicants with Microsoft Fabric. The project facilitates the analysis of quantitative and qualitative data from the Immigration, Refugees and Citizenship Canada (IRCC) website, predicts scores for Canadian Permanent Resident draw scores, leverages on RAG architecture via a chatbot interface, and visualizes the data using PowerBI.

The data used for the project originated from the Immigration, Refugees and Citizenship Canada (IRCC)
https://www.canada.ca/en/immigration-refugees-citizenship

Project Repository URL

https://github.com/Pelumioluwa/Microsoft-Fabric-Hackathon/tree/main

Project video

Congratulations 💜

Congratulations and thank you all who joined us for “HackTogether: The Microsoft Fabric Global AI Hack”. Whether you got a prize or not, you’re all winners to us. You decided to learn a new skill and made a deliberate effort towards it!

We’d love to hear what our readers think in the comments, let us know if there is a particular solution that stood out to you how you think it could provide value to end-users! If you would like to continue the discussion of this event, check out the Hack Together discussion board on the Microsoft Fabric Community site!

We understand that many of you found your way to this Hackathon as a means of flexing your skills to prepare for the DP-600 Fabric Analytics Engineer Certification.  We just announced a Microsoft AI Skills Challenge that will be launching on March 16, with pre-registration available today that will allow you to receive a FREE Microsoft Certification Exam voucher by completing a series of modules (including a specific Microsoft Fabric challenge!) 

Details are available on my LinkedIn (feel free to give me a follow if you’d like to say in touch!)

Cheers!

Related blog posts

Announcing the winners of “HackTogether: The Microsoft Fabric Global AI Hack”

April 23, 2024 by Misha Desai

At the recent Fabric Conference, we announced that both code-first automated machine learning (AutoML) and hyperparameter tuning are now in Public Preview, a key step in making machine learning more complete and widely accessible in the Fabric Data Science. Our system seamlessly integrates the open-source Fast Library for Automated Machine Learning & Tuning (FLAML), offering … Continue reading “Introducing Code-First AutoML and Hyperparameter Tuning: Now in Public Preview for Fabric Data Science”

April 18, 2024 by Santhosh Kumar Ravindran

We are excited to announce a new feature which has been a long ask from Synapse Spark customers, Optimistic Job Admission for Spark in Microsoft Fabric.This feature brings in more flexibility to optimize for concurrency usage (in some cases ~12X increase) and prevents job starvation. This job admission approach aims to reduce the frequency of … Continue reading “Introducing Optimistic Job Admission for Fabric Spark”