About Hackathons @ Slalom
Contact the team at email@example.com
Congratulations to Team Polly and Team Project Blue Cat!!
On Thursday, August 24, the top teams presented their projects to a panel of judges at an event that concluded the AI Hackathon. The final event had over 150 people in attendance with several clients and partners represented. (More details at the bottom of this email.) Our distinguished judges included Brad Jackson, Troy Johnson, Tom Chew, Alison Minnick, Luanne Pavco, and Justin Odenbach. Thank you to our judging panel for actively engaging the teams and making the tough decision on who would win. included two published skills and sutr.io, an open source tool to develop Alexa voice assets quickly.
After intense deliberations, the judges ultimately decided to award two teams the Slalom Hackathon Cup, previously held in Houston by the winners of the HoloLens Hackathon. Congratulations to team Polly from XM Seattle and team Project Blue Cat from Seattle!
Thank you to all teams for sharing your projects. You can see the deck presented at the event here, which can be shared with clients who have an NDA with Slalom.
Polly: Project Polly - XM Seattle
Facial Recognition And Tracking With Parrot AR
Tristan Nyyssela (lead), Josh Ruoff, John Schuldt, Shuan Hamman, Ian Burns
The Polly team from XM Seattle used open source projects, a Parrot AR 2.0 Drone and a DJI Mavic Pro Drone, and Azure Cognitive Services to develop a drone that can recognize a subject and track their movements. The drone is designed to keep the subject in frame at all times and follow that subject as needed. The team was able to use OSS projects as a starting point and thanks to a wonderful community of developers and droners the project was relatively straight forward. The purpose of the project was to bring to life a technology that could help solve many current issues. Things like finding a given person for search and rescue, helping police find suspects and fugitives, or even securing an area against specific individuals. The first phase of the project was written in Node.js for a Parrot AR drone using AWS services. The team was limited by the hardware (the Parrot Drone) mainly, and when given the chance to upgrade said hardware they jumped at it. The team made a change over to the DJI as well as a change to Azure services, this presented a few technical issues during our transition, including the need to port the project to android. The project is now written in Java for android, using OpenCV, DJI SDK, Azure Face API, and a bit of homebrewed knowhow. For our MVP, we use Azure Face API to create and train a person group with the specific subject we would like the drone to search for. After the drone is launched it will fly in a specified pattern looking for its trained subject. After the drone finds said subject, it will begin actively tracking the subject to keep them in view. We can then use the GPS coordinates of the drone to find the location of the subject.
Project Blue Cat - Seattle
Google Cloud Machine Learning
Fighting Human Trafficking with Artificial Intelligence (AI)
Gretchen Peri (Lead), Nick Drochak (PM), Colby Voorhees, Sue Grinius-Hill, Clinton Donough, Scott Macha, Matthew Netkow, Nate Holcomb, Daniel Spurling, Jim Bergh, Ryan Chapple, Tom Wagner, Chris Ottino, and Maggie Hamilton
Project Blue Cat worked to combat human trafficking using artificial intelligence by analyzing images and text to cluster online advertisements of commercial sex (using massage ads as a proxy). Their goal was to discover and disrupt the organized crime networks propagating human trafficking. Human trafficking is a $32 billion-a-year industry affecting 2-4 million women, children, and men who are sold every year for forced labor or commercial sex. Half of them are children. There are 100s of websites advertising commercial sex online including Backpage.com and Craigslist. Project Blue Cat is analyzing online ads for similarities in structure, images, and content to group ads into correlated clusters. This enables law enforcement and prosecutors to focus on groups of ads that display similar characteristics and potentially a trafficking network as well as ads posted across multiple cities. The team built a Proof-of-Concept that currently does the following: scrapes online ad content from Backpage.com, analyzes images using Google Cloud Vision API, analyzes text using Google Natural Language Processing (NLP) API and Data Loss Prevention (DLP) API, and exposes results through a web UI for interactive search on key fields and identifies probable clusters of organized crime networks involved in human trafficking.
Which Partners and Client attended?
- Arthur J. Gallagher & Co (AJG)
- Beam Global Spirits & Wine, Inc.
- Cushman & Wakefield
- General Growth Properties, Inc. (GGP)
- Haemonetics Coporation
- JPMorgan Chase & Co. (JPMC)
- US Foods
- Zurich North America
Which Markets Participated?
- Los Angeles
- New York
- Orange County / San Diego
- XM Boston
- XM Chicago
- XM Houston
- XM Seattle
- XM Toronto
Who Runs the XM Hackathon program?
The XM Hackathon series is sponsored by Mike Cowden, Ian Cook, and Greg Martin out of the XM Delivery Center in Seattle. The Hackathon program is lead by a committee of cross-functional members consisting:
- Jeremiah Dangler (XM Seattle - Engineering) - Program Director
- Adam Limoges (XM Seattle - Alliances) - Alliance Lead
- Jon Busby (Chicago - Business Advisory Services) - Business Lead
- Jeff Northcutt (Seattle - Business Leadership) - Client Lead
- Harpreet Sandhu (XM Seattle - Operations) - Operations Lead
- Kim Kwock (XM Seattle - Marketing Studio) - Marketing Lead
Contact the team @ firstname.lastname@example.org
2017 AI Hackathon.
Today we continue the XM Hackathon series with the 2017 Artificial Intelligence (AI) Hackathon that will focus on five AI platforms: Amazon AI, Azure Cognitive Services, Google Cloud Machine Learning Platform, IBM Watson, and Salesforce Einstein. These platforms offer tools to perform machine learning, text-to-speech, speech-to-text, image recognition, chat bots, and more.
What follows is the framework and rules for the AI Hackathon. Please email email@example.com for any questions. Please check back on a regular basis as we update with more content.
It's almost time to submit your project. Teams must complete this survey.
- April – Get Started
- Form your team
- Select platform(s)
- Develop Idea
- Start POC / Prototype
- May – Continue Development
- Sign your team up (link coming)
- Iterate over your POC / Prototype
- June – Continue Development
- Iterate over your POC / Prototype
- July 7th – Submit Your Project
- Team Submission Due
- 10 to 15 slide deck
- 4 - 5 minute video
- Source Code
- Teams must complete this survey
- Team Submission Due
- July 14th – Top teams Selected
- Top teams announced and invite to the finals
- Teams arrange flights
- August 23rd – 25th Finals
- Location: Chicago
- Three-day event
- Markets must cover all time and expenses for its team.
- You must work with your Project Lead and People Manager to ensure your client delivery is not affected
- Your project will be inclusive of at least one of five AI platforms: Amazon AI, Azure Cognitive Services, Google Cloud Machine Learning, IBM Watson, Salesforce Einstein
- Teams will pick a platform(s) and develop something cool, innovative, useful, or all of the above!
- It’s encouraged that teams be cross-practice (Engineering, XD, IM&A, etc.)
- Please feel free to form teams amongst all markets, all locations as you see fit. All project work must be exclusive to Slalom employees, however, you are able to ideate directly with a Slalom client.
- Your team should select a team name and a Slalom home office
- Recommended team size is 4-8 members.
- Your team, should you make it to the Finals, will be expected to present in front of Slalom leadership in Chicago
- Innovation and Creativity
- How creative was the idea? Why does the idea matter?
- Was the approach of the idea unique? What was the potential of the idea?
- Was an MVP declared? Did a product roadmap exist? Was there a clear vision?
- Technical Difficulty
- What technical breakthroughs were relevant? What was the complexity of the design?
- Any particularly interesting frameworks or integrations that were used?
- Was there any unique usage of technology or services?
- Did the team implement a POC or demo? How was it? Was the demo interactive?
- How was the work distributed on the team? How were different skill sets used?
- Did the team overcome difficult obstacles?
- Applicability to Business Objectives
- What was the core problem being addressed? What market opportunities exist?
- Were unique and key product features explained in sufficient detail?
- Were unique AI capabilities used in a creative or different way?
- Did the team connect their idea to potential clients or industries?
- Did the team connect Slalom to AI in a unique way?
- Did the team leverage the AI platform capabilities, at scale, in both creation & presentation?
- What type of project qualifies?
- We realize AI is a big term, and a big platform. We are not limiting your creative desires but your project MUST include elements of AI. Get creative! Maybe think of a client/industry/demographic this would be relevant to.
- What type of team should I have?
- Teams should be cross-functional (Business, IM&A, NGI, XD, CE, OE, Delivery, Engineering, etc.) to have the most well-rounded perspective in your project. Your team can be cross-location as well.
- Are their AI Platform costs (cloud/licensing) during the Hackathon?
- AI platform costs are at the discretion of the market. Of the platforms available, the request guidelines are as follow (insert link). If there are platform, licensing, or cloud costs, the burden falls on DC/Local Market. We suggest you try to be as frugal as possible with your project.
- What's included in the final presentation?
- We will ask (at minimum) prepare a presentation and video. (10 Page PPT + < 5 Min. Video, both required)
- How will your project be judged?
- Please see the 'Judging Criteria' slide
- Can we use an AI Platform outside of the five pre-selected?
- Yes, however one of the five AI Platforms MUST be primary in your project
- Can we use more than one AI platform?
- Yes, integrating platforms can produce interesting results
- Can we work with a client?
- Yes, however your submission may limit external exposure and PR for your team. Your team should only be composed of Slalom employees. You must avoid copyright infringement and IP issues. You may use a client to help ideate on your project.
- Can non-Slalomites participate in the Hackathon?
- The short answer is no, but you can work with clients to develop ideas and provide support. If non-Slalomites work on the project they may not be able to attend the Finals. You are allowed to collaborate with a client, but this may exclude your work from being shared externally due to client related IP.
- How do I sign up?
- To register you and your team for the 2017 AI Hackathon, please click here.
- Microsoft - Azure Sandbox (request form here)
- You must identify an Azure Ambassador to approve
- This Azure cost is billed back to your market
- AWS - AWS Innovation Lab (request form here)
- You must identify an AWS Ambassador to approve
- Google - We do not currently have ability to provide access
- SalesForce - We are a Salesforce Platinum partner and have access to test Beyond Core tools for testing up to 50,000 records. For more detail contact Mike Jortberg and join the Chatter group Salesforce AI Einstein here
- Contact Mike Jortberg for access to Slalom Salesforce test account
- IBM - We do not currently have ability to guide/provide access
What follows is a quick overview of what each platform provides, this may not include all service and apis provided.
- Amazon AI - https://aws.amazon.com/amazon-ai/
- AI Services
- Amazon Rekognition – image recognition service for image and facial analysis
- Amazon Polly – text-to-Speech service that uses deep learning to turn text into lifelike speech
- Amazon Lex – chat bot service for automatic speech recognition (ASR) for converting speech to text, and natural language understanding (NLU) to recognize the intent of the text
- AI Platforms
- Amazon Machin Learning – service that allows you to train custom machine learning models using your data
- Amazon EMR – managed Hadoop framework that makes it easy, fast, and cost-effective to process vast amounts of data
- Spark & SparkML – scalable machine learning algorithms
- AI Frameworks that support researcher and data scientist that want to build sophisticated and cutting-edge systems
- AI Services
- Microsoft Cognitive Services - https://www.microsoft.com/cognitive-services/en-us/apis
- Computer Vision API – distills actionable information from images
- Content Moderator – automatically moderate text, images, and videos
- Emotion API – personalize experiences with emotion recognition
- Face API – detect, identify, organize, and tag faces in photos
- Video API – analyze, edit, and process videos
- Bing Speech API – convert speech to text and back again, and understand intent
- Custom Speech Service – fine-tune speech recognition
- Speaker Recognition API – give your app the ability to know who’s talking
- Big Spell Check API – Detect and correct spelling mistakes
- Language Understanding Intelligent Service – teach your app to understand commands
- Linguistic Analysis API – easily parse complex text with language analysis
- Text Analytics API - detect sentiment, key phrases, topics, and language from your text
- Translator API – translate simple speech and text with a simple REST api
- Web Language Model API – leverage the power of language models trained on web-scale data
- Academic Knowledge API - explore relationships among academic papers, journals, and authors
- Entity Linking Intelligence Service - contextually extend knowledge of people, locations, and events
- Knowledge Exploration Service - add interactive search over structured data to your project
- QnA Maker - distill information into conversational, easy-to-navigate answers.
- Recommendations API - provide personalized product recommendations for your customers
- Bing Autosuggest API - give your app intelligent autosuggest options for searches
- Bing Image Search API - bring advanced image and metadata search to your app
- Bing News Search API - link your users to robust and timely news searches
- Bing Video Search API - trending videos, detailed metadata, and rich results
- Bing Web Search API - connect powerful search to your apps
- Google Cloud Machine Learning - https://cloud.google.com/products/machine-learning/
- Large Scale Machine Learning Service
- build sophisticated, large scale machine learning models that cover a broad set of scenarios from building sophisticated regression models to image classification.
- Job Search and Discovery
- highly intuitive job search that anticipates what job seekers are looking for and surfaces targeted recommendations that help them discover new opportunities
- Video Analysis
- makes videos searchable and discoverable by extracting metadata, identifying key nouns, and annotating the content of the video
- Image Analysis
- enables you to understand the content of an image by encapsulating powerful machine learning models in an easy to use REST API
- Speech Recognition
- enables you to convert audio to text by applying neural network models in an easy to use API.
- Text Analysis
- reveals the structure and meaning of text by offering powerful machine learning models in an easy to use REST AP
- Fast, Dynamic Translation
- provides a simple programmatic interface for translating an arbitrary string into any supported language
- Large Scale Machine Learning Service
- IBM Watson - https://www.ibm.com/watson/developercloud/services-catalog.html
- Conversation - build chatbots that understand natural language and deploy them on messaging platforms and websites, on any device
- Document Conversion - transform documents into different formats
- Language Translator - translate content into multiple languages
- Natural Language Classifier - interpret and classify natural language with confidence
- Natural Language Understanding - natural language processing for advanced text analysis
- Personality Insights - understand personality characteristics, needs, and values in written text
- Retrieve and Rank - enhance information retrieval with machine learning
- Tone Analyzer - understand tone and style in written text
- Speech to Text - convert human voice into written word
- Text to Speech - enable computers to speak like humans
- Visual Recognition - understand image content
- Data Insights
- Discovery - rapidly build a cognitive search and content analytics engine
- Discovery News - find insights from news and blog content enriched by Watson for highly targeted search and trend analysis
- Tradeoff Analytics - make better choices with a full view of your data
- Salesforce Einstein - https://www.salesforce.com/products/einstein/overview/
- HipChat Room - Hackathon: Artificial Intelligence
- AWS Public Datasets - https://aws.amazon.com/public-datasets/
- Public Datasets - https://github.com/caesar0301/awesome-public-datasets
- City of Chicago Datasets - https://data.cityofchicago.org/
- Uber Dataset - https://movement.uber.com/cities
- Slalom Salesforce Sandbox - Contact Mike Jortberg
- IBM Watson Overview PPT deck and video
- Envision - Artificial Intelligence
- Google Specific
- "Machine Learning APIs by Example (Next '17)" (YouTube) - overview of our Machine Learning APIs (Speech, Vision, Translate, etc).
- "TensorFlow and Deep Learning without a PhD (Next '17)" (YouTube - Part 1 & Part 2)
- "Introduction to Google Cloud Machine Learning (Next '17)" (YouTube)
- "Lifecycle of a machine learning model (Next '17)" (YouTube)
- "BigQuery and Cloud Machine Learning: advancing neural network predictions (Next'17)" (YouTube)
- "Introduction to big data: tools that deliver deep insights" (YouTube)
- "Preventing Overfishing with Machine Learning and Big Data Analytics (Next'17)" (YouTube)
- "Auto-awesome: advanced data science on Google Cloud Platform (Next'17) (YouTube)
- XM Boston - Justin Lee (firstname.lastname@example.org)
- XM Chicago - Tony Bergeron (email@example.com)
- XM Houston - Victor Aparicio (firstname.lastname@example.org)
- XM Seattle - Jeremiah Dangler (email@example.com)
- XM Toronto - MJ Alwajeeh (firstname.lastname@example.org)
- Dallas - Patrick Swain (email@example.com) & Tracey Kelly (firstname.lastname@example.org)
- New York - Lee McOmber (email@example.com)
- St. Louis - Nate Johnson (firstname.lastname@example.org)
- Atlanta - Todd Heil (email@example.com) & Jeremy Lizza (firstname.lastname@example.org)
- OC/SD - Seyed Ketabchi (email@example.com)
- Chicago - Jon Busby (firstname.lastname@example.org)
- Los Angeles - Alen Zograbyan (email@example.com)
- Seattle - Nick Drochak (firstname.lastname@example.org)
- Minneapolis - Kari Krautbauer (email@example.com)
To register you and your team for the 2017 AI Hackathon, please click here.
- Updated Salesforce Licensing information
- Added Los Angeles
- Added Hackathon Team
- FAQ added: Can non-Slalomites participate in the Hackathon?
- Added Seattle
- Resources add: City of Chicago Datasets and Uber Datasets
- Added Minneapolis
- Added Slalom Salesforce Sandbox information
- Added registration link
- Added IBM Watson Overview PPT and video
- Added Envision - Artificial Intelligence event link
- Added additional Google Resources
- Added link for Team Submissions
- Added finals recap