Recent Use-cases of Data Science & Analytics, Machine Learning, Deep Learning and Automation

Source: Deep Learning on Medium

Recent Use-cases of Data Science & Analytics, Machine Learning, Deep Learning and Automation

The terms Analytics, Data Science, Machine learning, Deep learning and Automation are becoming more prominent. The question really is how are businesses using these technologies, how are they using it drive revenue, improve product and increase efficiency. Here, we would be discussing some use-cases of these technologies in various economic segments.

A. Chain Restaurants

Chain Restaurants use Data Science & Analytics to analyze the impact of a new menu on the market, they do it to figure whether the new menu can drive enough sales to offset the cost of marketing the new menu. The effect of the new menu on the sampled locations are analyzed using A/B test and a recommendation is given as to whether the Chain restaurant should launch this new menu.

Chain Restaurants also glean insights from looking at their purchase data, Once you know that people who are more likely to buy ‘Menu A’ might also like item ‘Menu B,’ you’re also getting a lot of insight from the consumer on not only a promotional standpoint but also on a bundling standpoint and that information can help you adjust the menu.

McDonald’s corp recently acquired Dynamic Yield, an Israel-based technology company specializing in machine learning that provides product suggestions to e-commerce shoppers based on their orders and other variables.

B. Manufacturing / Production Industries

Manufacturer use Analytics to forecast inventory. Manufacturers use historical production data to know how many goods (such as generators, tricycles) they’ll need to produce over the next six months to meet expected demand. Since the outcome the manufacturer wants to predict is a number, then the target variable is numeric. Therefore, a numeric or regression model can be used to solve this problem.

Companies like General Electric, with its more than 500 factories and thousands of factories in its supply chain, are taking advantage of data analytics at the enterprise level. They use predictive applications to forecast inventory demand. Benefits of such systems include overall improved quality with fewer downtime operations, faster cycle times, and greater savings realized in labor costs and improved operational efficiencies across the enterprise and the globe.

C. Fast Moving Consumable Goods business

Fast Moving Consumer Goods industries such as (Processed foods, Prepared meals, Beverages, Medicines) use Data Science & Analytics in a variety of ways.

Pizza chains use sales data from their existing stores and respective demographic data around those stores to predict how many pizzas they’ll sell at their new store location. Since the outcome that the Pizza is trying to predict is the number of pizzas, then the target variable is numeric and they would use a numeric or regression model to solve this problem.
This also applies to Ice cream vendors who sells a different amount of ice cream every week. they puts in orders for ice cream containers once a week, but would like to know how much to order so they have enough ice cream to sell, but doesn’t spend too much on keeping excess inventory frozen.

Domino Pizza captures data on all its channels — text message, Twitter, Pebble, Android, Amazon Echo — to name just a fraction — and feed into the its Domino’s Information Management Framework. There it’s combined with data from third party sources such as the United States Postal Service as well as geocode information, demographic and competitor data, to allow in depth analytics and customer segmentation.

D. Credit/Risk-Management Companies

A credit organizations provides loans to small businesses. Once a pre-loan application is approved, the applicant has 90 days to apply for the actual loan. The financial organization use Data Science & Analytics to predict how many loan applications they have to process from their pool of pre-loan applicants.
Credit companies / Risk Management Department at Banks also use historical data of their clients to predict whether a new customer will default on a loan, always pay on time, or sometimes pay. Since the outcome the bank is trying to predict is a category that the new customer will fall into, non-numeric or classification model is used to solve this problem.

Amazon loans billions of dollars to small businesses reselling on its platform. Machine learning is used to identify borrowers who are low credit risks based on their inventory turnover and profitability. Amazon relies completely on machine learning, so no humans are involved, not even with filling out an application, and it offers unsolicited loans on “take it or leave it” terms.

E. Advertising/Marketing Companies

Marketing organization use analytics to predict whether someone is likely to redeem a coupon as they would like to minimize costs and only send coupons to people who are likely to use them.

Influans uses big data architecture and AI algorithms to create more targeted and personalized marketing for brands and retailers.

F. Media Companies, Customer Service

AI is now enabling businesses to better manage and find value in their enterprise video and audio assets.

Deep learning technology empowers businesses to understand and optimize video content libraries with advanced metadata enrichment and previously untapped insights. From heightening engagement and increasing discoverability to automating closed captions and furthering inclusivity.

Deep learning is also enabling Media companies or Customer relations industries transcribed recordings/calls in real-time or near real-time, and the resulting transcripts can be used for advanced analytics. These text transcripts can be stored while the high-quality uncompressed audio files are able to now be deleted and don’t need to be stored. The ability of companies to provide real-time access to this data has also required advances in how the data is stored and processed.”

Gridspace uses deep learning networks to drive sophisticated speech recognition systems. Their networks begin with raw audio and build up to topic, keyword and speaker identification. Their Gridspace Memo software is designed to identify speakers, keywords, critical moments, and time-spent-talking, along with providing group take-aways from conference calls. Gridspace Sift provides similar information about customer service and contact calls.

G. Recruitment Agencies

Recruitment Agencies are using Data Science to predict Employee eligibility using decision tree models whose parameters have been optimized by GridSearchcv..

Big data and AI tools are increasingly being offered by Human Resource business like LinkedIn to sift through candidates’ profiles and identify the most suitable people for a position. Which is just as well considering that 52% of talent acquisition leaders state the most difficult part of recruitment is identifying the right people from a large pool of applicants.

JetBlue Airlines gives us one great example of data analytics being used to find the most suitable candidates. Previously, the company had focused on ‘niceness’ as the most important attribute for flight attendants. Then, after carrying out some customer data analysis with the Wharton Business School, JetBlue was interested to find that, in the eyes of their customers, being helpful is actually more important than being nice — and can even make up for people being not so nice. The company was then able to use this information to narrow down candidates more effectively.

H. Web & Mobile apps Development

Business now deploy Data Science / Machine learning models into Web and Mobile applications.

Web apps — These includes Interactive analytical dashboard with predictive capabilities like a customer/ employee churn prediction dashboard, a stock price prediction dashboard.

Mobile apps — These include machine learning apps like Chatbots or virtual assistants, built on conversational data, can predict a suitable answer to a questions, For example Siri and Google assistant virtual assistant uses Natural Language Processing, Its a machine model for converting speech to text.

Some real use case of machine learning integration into mobile apps include Google smart compose, the smart compose feature is powered by machine learning and will offer suggestions as you type. It involves an LSTM model that will predict the next likely text to a typed sentence. It enables one write emails faster. Also Snapchat app overlays emojis on pictures and videos using machine learning model to detect faces and locate the facial features.

I. Retail/ Finance

Retails businesses use Data Science and analytics to answer questions like
• Product: Is the business carrying the right goods at the right time (Customer-centric localised assortments)? What is behind sale and margin fluctuations? What is the current stock situation and is an out-of-stock likely?
•Pricing: What are the pricing sweet spots? How do price adjustments affect sell-through? What tactics can be employed to manage markdowns and control margin erosion?
• Promotions: What products do customers typically buy together? Can promotion spend be optimised?
• Product placement: Are products meeting sales targets? What impact can repositioning have on specific items? What is the yield effectiveness of shelf, aisle and end-cap placements?
• In-store: Are stores staffed and managed optimally? Do store associates have real-time customer insights? Are customers provided with personalised experiences?

Nike said it is acquiring Celect, a retail analytics company, to build out its data science skills for its Consumer Direct Offense plan. Celect’s cloud analytics platform enables retailers to optimize inventory across channels. The platform enables efficient order management, allocation, planning and In-Season decision making.

J. Online Businesses (Digital Products/Goods)

Online Businesses use Data science and machine learning techniques to create simple algorithms, which analyze and filter user’s activity in order to suggest him the most relevant and accurate items. Such recommendation engines show the items that might interest the user, even before he searched for it himself.

To build a recommendation engine, data specialists analyze and process a lot of information, identify customer profiles, and capture data showing their interactions to avoid repeating offers. For example there are business using recommender system to recommend digital products such as video content, news articles, data bundle/ airtime options from ISPs and Telecoms, loan services, consumable products on e-commerce websites.

YouTube and Netflix use recommender system to suggest video content to its customers based on by how high they rate a content, based on content liked by people similar to them, based on how much time was spent viewing a content and based on generally popular content.

K. Operations

The large volume of invoice processing has repetitive manual tasks which can result in delayed and incorrect payments.Invoice processing has many challenges such as invoices with formats, require consolidated data from various sources into the single financial database system etc.

Robotic Process Automation (RPA) automatically processes invoices once received. RPA can automate the data input, reconciliation error, and even it can process certain decision-making required for invoice processing, which minimizes the need for human intervention.The paper-based invoice format can automate using OCR (Optical Character Recognition)

In the United States, Cleveland-based KeyBank is looking for efficiency gains, and it recently partnered with Billtrust to automate invoice delivery and accounts receivable services.

Through electronic invoicing on a cloud-based platform, building on Billtrust’s Quantum Payment Cycle Management, the bank can now incorporate RPA principles into activities involved in processing of accounts receivable such as generating invoices, maintaining records of payments due and payments received, extending credit, and performing accounting functions. Specifically, the RPA technology at play eliminates the need for employees to carry out repetitive manual tasks.