dev hackathon - 3rd editionhackathon.retail.infor.com/dev/hackathon - 3rd edition.pdf ·...
TRANSCRIPT
October 2016
1Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Dev Hackathon - 3rd editionApril 2018
October 2016
2Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Plan
I. Welcome to INFOR!II. Introduction to the Retail Data Model
III. Data Partitioning ChallengeIV. Your MissionV. Evaluation Criteria
October 2016
3Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential 3Copyright © 2015. Infor. All Rights Reserved. www.infor.comInfor Confidential
Welcome to INFOR!
THE BIGGEST COMPANY YOU’VE NEVER HEARD OF
Infor’s global presence
3rd largestbusiness applications
company
14,000employees
$100Btrade in commerce
cloud
25,000 suppliers in
commerce cloud
30 banks in
commerce cloud
1,248+go lives
73,000customers globally
200+countries & territories
$2.8Brevenue
4,500+cloud customers
globally
October 2016
5Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Welcome to INFOR!
• INFOR RETAIL provides cloud native, predictive and prescriptive analytics solutions for retail. Our Solutions harness big data and advanced analytics to solve the toughest problems in merchandising and supply chain.
MERCHANDISING PLANNING
ASSORTMENT & CATEGORY
MANAGEMENT
DEMAND FORECASTING
PRICE OPTIMIZATION
SUPPLY CHAIN OPTIMIZATION
A Seamless & configurable end to end
planning suite
Localized, optimized assortments that
drive higher sales & margins
Deeper insights into consumer
demand driving up to 50% improved
forecasts
Optimized price, promotion &
markdowns linked to merchandising &
supply chain
Optimizes all network locations & flows
globally – 20% inventory reduction
October 2016
6Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
● A yearly software engineering hackathon that addresses real-world engineering problems.
● This year , we have 2 challenges :
○ A development challenge. ○ A data science challenge.
Welcome to INFOR!
Welcome to the THIRD edition of Infor Tunisia Hackathon
October 2016
7Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential 7Copyright © 2015. Infor. All Rights Reserved. www.infor.comInfor Confidential
Introduction to the Retail Data Model
October 2016
8Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Introduction to the Retail Data Model
• Let’s imagine that you landed a job with a big American retailer (sweet!).• This retailer has many stores spread across multiple cities. • Many items are sold in each one of those stores. • Items are grouped into classes to facilitate customer browsing and
internal bookkeeping.
Location CalendarProduct
October 2016
9Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Introduction to the Retail Data Model - Dimensions
DimensionlevelslevelsLevels
Product
October 2016
10Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Introduction to the Retail Data Model - Hierarchies
Smartphones
iPhones Samsung phones
Samsung S5Samsung S4iPhone 8iPhone 7
Class level
Subclass level
Item level
October 2016
11Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Introduction to the Retail Data Model - Metrics
● Our retailer keeps track of different metrics about its business including:
○ The number of units sold of each of the items he sells every day in each store.
■ units [item, store, day] = value
○ The profit generated from selling an item during each day in each store.
■ profit [item, store, day] = value
October 2016
12Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Introduction to the Retail Data Model - Metrics
Metric Keys Metric Values
Metric Name
October 2016
13Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential 13Copyright © 2015. Infor. All Rights Reserved. www.infor.comInfor Confidential
Data Partitioning Challenge
October 2016
14Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Data Partitioning Challenge
• What is partitioning?– It’s a technique for physically dividing the data into
smaller pieces according to some predefined rules.
• Why partitioning?– Processing of large amounts of data (BIG Data)– Parallelizing data analysis– Improving maintenance, performance of our systems
October 2016
15Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Data Partitioning Challenge
HUGE Sales File
Partition by:1. Class level of
the Product Dimension
2. City level of the Location Dimension
Class level valuein Product Dimension
City level valuein Location Dimension
Sales records of the class “Smartphones”
in the city “ATL”
October 2016
16Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential 16Copyright © 2015. Infor. All Rights Reserved. www.infor.comInfor Confidential
Your Mission
October 2016
17Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential 17Copyright © 2015. Infor. All Rights Reserved. www.infor.comInfor Confidential
Evaluation Criteria
October 2016
18Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor Confidential
Evaluation Criteria
• Solution correctness• Passed levels • Performance• Code quality
October 2016
19Copyright © 2016. Infor. All Rights Reserved. www.infor.comInfor ConfidentialINFDTP1482386-en-US-0816-3