Media Case Studies

Issues & Objectives

• One of world’s largest primary cocoa manufacturer having 60+ plants across the globe
• Procures cocoa beans from across the globe and supplies cocoa butter, liquor and cakes to all large brands across the world
• Objective was to minimize the cost of procuring & storing cocoa beans While ensuring that sales forecasts, and sales orders, as they get confirmed are met

Results

• Reduced stock from 24 months to 7 months while always ensuring at least 3 months’ buffer for procurement time of 1 – 3 months
• Consistently met quality requirements
• Reduced unusable old beans stock to Zero

Project information

Techniques

Mathematical Optimization

Client

Large Cocoa Manufacturer

Industries

Supply Chain & Retail

Location

Singapore

Challenges

• Large number of recipes
• Converting monthly to daily production plan imposes practical limitations

Solution

• Propose the following while meeting quality requirements

• Annual Procurement Plan: minimize cost of produce, transit and holding and meeting sales forecast
• Monthly Production Plan: meet monthly demands of cake, butter, liquor and powder
• Daily Production Plan: meet daily sales demand and ensure maximum utilization of lines

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• The trucking company was facing challenges in agreeing on right freight price with truck operators.
• Only a few skilled agents with knowledge of certain lanes could negotiate successfully
• Suggest appropriate fare to be offered on load-boards for full truck loads (FTL) freight on a specific lane (Origin-Destination) for a particular date, equipment & load features, in the continental US market

Results

• A combination of very accurate localized models and broader hub-based models gave ~95% accuracy

Project information

Techniques

Forecasting

Client

Large Trucking Company

Industries

Supply Chain & Retail

Location

US

Challenges

• Very sparse data: 70,000 loads (over 3 years), spread over 70 lanes and 42 equipment to answer a problem for any lane across the continent
• Apparently erratic prices: e.g. OH-TX lanes (750-1200 mi) are priced same as IN-PA (500-600 mi), for same equipment!
• Prohibitive price of historic data from Truckstop or DAT meant only generic third-party data like Fuel prices & CASS index

Solution

• PDFs need to be OCR’d to load texts in it
• Ensemble modelling – A very large number of models were developed
• Solution deployed using MLFlow, integrated with DataLake and transaction system

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• A leading manufacturer of hair care products in Bangladesh wanted to plan marketing campaign and distribution strategy with accurate timing and execution at micro level
• The company needed to achieve this through anticipation and planning for cycles and risks
• Project objective was to develop predictive demand model for its leading brand and one newly launched brand 3.5 months in advance
• The predicted change in consumer demand would enable them to take timely actions

Benefits

• Models predicted future sales with at least 90% accuracy
• Predictions using various scenarios of ad spend and distribution helped the company plan strategy and to be ahead of competition

Project information

Techniques

Mathematical Optimization

Client

Large Trucking Company

Industries

Supply Chain & Retail

Location

US

Challenges

• Collection of macro economic variables from various sites
• Alignment of variables from differing time periods
• Alignment of retail audit and household panel data

Solution

• Forecast macro economic variables using time series models (ARIMA, Holt winter, Exponential smoothing)
• Predict disposable income from macro economic variables
• Forecast category offtake from disposable income and past offtake taking into account seasonality, festival, distribution
• Predict brand sales from category offtake and other controllable factors

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• To schedule content (movies, shows etc.) for linear TV which optimizes viewership balancing content rights, various campaigns, audience fatigue and airing rules.

Benefits

• Can be used for scheduling any content – music, shows (sitcoms) and movies
• Improved content rights utilization
• Improved audience targetting
• Controlled release of content to minimize viewer fatigue
• Strict adherence of airing rules
• Flexibility to plan campaigns based on genre, cast, theme etc

Project information

Techniques

Mathematical Optimization

Client

Leading Broadcasters in India

Industries

Media

Location

India

Solution

• Interface allowing multiple users to collaboratively input campaigns
• Pre-specify certain airings
• Take inputs over APIs or database view for airing rights, viewership history etc
• Provide airing rules, specify primary TGs for timebands
• Generate programming schedule with tags
• Users can prepare multiple competing schedules, run comparisons and decide best schedule well in advance

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• One of world’s largest mining corporation having a network of 9 mines
• Approximately $ 400 million working capital used to manage spares inventory of about 182K SKUs
• Stock position monitored by the MRP system every ½ hour & orders are placed
• Objective was to propose improvement to spares inventory management at each storage location to minimize locked up working capital while meeting SLAs

Benefits

• 20% savings in working capital ~ $ 80 million
• The existing MRP database was updated with new ROP & EOQ

Project information

Techniques

Mathematical Optimization

Client

Largest Mining Corporation

Industries

Supply Chain & Retail

Location

Singapore

Challenges

• Large number of SKUs ~182K
• Demand variations

Solution

• Derived inventory policy at the network level from ROP and EOQ
• Identified 9 different models for distributions of demand volume
• Developed an ML decision tree for choosing appropriate model based on value & velocity
• Computed EOQ using selected model

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

Transport regulatory authority in Singapore commissioned a system to

• Automatically discover wrong fare incidents and flag commuter’s cards affected by wrong fare charging; and
• Detect emerging fault trends in fare collection equipment so that corrective action could be taken in a timely manner

Challenges

• Large data 15 million transactions per day, which translates to more than 5 billion historical transactions in a year needs to be processed to identify fault patterns and trends
• The data consisted of financial, operations, transit and events data of buses

Project information

Techniques

Forecasting

Client

Transportation Authority

Industries

Public Sector & Education

Location

Singapore

Solution

• Data Storage: Hadoop and MySQL
• Query Tools: Hive and SQL
• Algorithms: rmr (Parallel versions of R) and Java
• Reporting and Dashboards: Pentaho

Results

Robust solution in use for over 2 years allows pro-active rather than reactive maintenance

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• Long term infrastructure planning – over a 52 year horizon
• Determine the sequence in which sub-sea wells should be drilled to maximize profit

Project information

Techniques

Mathematical Optimization

Client

–

Industries

Supply Chain & Retail

Location

–

Solution

• Given 140 polygons (indicated by lat/long) which ones should be drilled?
• When should each polygon be drilled?
• Well platforms are needed to support the drilling of wells
• What is the number of well platforms required? What capacity should each have? When should we commission each platform?
• Hydrocarbon flow from wells will be stored and processed at production platforms
• What is the number of production platforms required? What capacity should each have? When should we commission each platform?
• It is necessary to make these choices together and not sequentially
• Number of rigs available
• Number of polygons to be drilled in any period limited by available capacity of well platforms
• Well flow rate in any period limited by available production capacity
• Problem modelled and solved using complex optimization techniques
• Provides a critical strategic planning tool for senior management

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• A Singapore based company provide multi country mobile platform for distributed sales representatives who gets updated information on demand forecast, recommendation and target sale
• They wanted to build appropriate models for forecast
• All output were to be pre processed in nightly batch run and saved in a centralized database
• A customized software for managerial decision making was also needed

Benefits

• Batch run for a dataset of 60K transaction take less than 10 minutes producing multiple output tables
• Experiment with customer segments and view a particular subset for any discount/promotion
• View the position of customers and the recommendation to be made
• Review the profile of DSR and extent of target achievement
• Employ Various methods and visualize actual vs forecast

Project information

Techniques

Forecasting

Client

SaaS Provider

Industries

Supply Chain & Retail

Location

Singapore

Challenges

• High attrition of DSRs made it hard to collate a time series sales data
• Customer base changes between transition from one DSR to another
• Intermittent sales data for about 30% of customers
• Discontinued or new product SKUs with short history of sales data

Solution

• Software developed in R Shiny
• K-means and hierarchical clustering and time series forecasting methods were used
• Batch code is developed in R with input and output link to client database

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• Forecast passenger and revenue for major O&D (Origin & Destination)/POS (Point of Sale) combinations for a large East African Airline
• Short term O&D forecasts for every flight date up to 90 days in the future to be generated everyday
• Long term rolling forecasts up to 5-10 years to be generated quarterly

Methodology

• Linear Regression
• ARIMA/ARIMAX
• Neural Networks
• Etc.

Project information

Techniques

Mathematical Optimization

Client

Budget Airlines

Industries

Supply Chain & Retail

Location

Africa

Data

Short term forecasts based on:

• Current bookings
• Historical bookings
• Seasonality
• DOW (Day-of-Week)
• Etc.

Long term forecasts based on:

• GDP
• Population growth at origin
• Population growth at destination
• Employment growth at origin
• Employment growth at destination
• Etc.

Solution

• O&D forecasting is very challenging because of the small numbers involved
• Good accuracies obtained

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• Generate in real time (typically a few minutes) the daily spot allocation plan which determines the program/breaks in which each spot will be aired

Benefits

• Maximizes revenue ( 2-4% incremental gain)
• Automates the spot allocation process
• Respects FCT Caps
• All allocation rules such as cap on number of ads for the same brand in a program are satisfied
• Checks that all deal conditions are satisfied while allocating ads

Project information

Techniques

Mathematical Optimization

Client

Leading TV Broadcaster

Industries

Media

Location

India

Solution

Designed and developed a software with following features

• Assured allocation at spot, brand, advertiser, deal level
• Even distribution of spots of different brands, products, clients when the rates are same
• Long term even distribution of spots across day-parts from each deal time-band

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• An end to end system to automate and optimize advertisement inventory planning for media resulting in additional revenue gain
• The solution focuses on scheduling of commercials, creation of proposals, planning of ad inventory and post evaluation

Benefits

• Maximizes revenue ( 2-4% incremental gain)
• Suggests profitable deals in real time
• Ensures servicing of deals
• Saves premium inventory for selling at higher price
• Right pricing through visibility of inventory
• Saves time and cost
• Reduces people risk

Project information

Techniques

Mathematical Optimization

Client

Leading Broadcaster in India

Industries

Media

Location

India

Modules

• Proposal Builder – customizes rates and inventory commitment to satisfy advertiser/agency requirements while maximizing broadcaster’s margins. It can take care of both CPRP and ER based deals & simultaneously address the requirements of large advertisers and the scatter market
• Inventory Visualizer – Inventory Visualizer derives insights for sales strategy planning from data. It provides clear visual presentation of current and historical data, insights for right pricing and full visibility of inventory, consumption and availability
• Ad-Spot Optimizer – generates in real time (typically a few minutes) the daily spot allocation plan which determines the program/breaks in which each spot will be aired
• Demand Forecaster – The demand Forecaster forecasts demand from forecasters for day-parts and programs taking into account  factors such as booking history, current bookings, channel grp, seasonality, festivals and economic indicators
• Post Eval – tracks the performance of marketing campaigns by mapping spot ratings to as-run logs
• Make Goods – suggests appropriate spots for make good to compensate for dropped ROs or unmet GRP targets while maximizing revenue

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• Loans by a large financial company in India to real estate developers are repaid in a mix of cash and inventory
• The projects under development will only be ready for launch (sale to buyers) a number of years after the loan is taken
• The financial company therefore requires prediction of property prices at time of launch and a number of years post launch
• They commissioned Smart to build customized software for such price analytics

Benefits

• Property valuation at launch as well as comparison with competition in few clicks
• View relative position of the project in the same locality
• Valuation in next 5 years post launch is instantaneous

Project information

Techniques

Forecasting

Client

Large financial company in India

Industries

Supply Chain & Retail

Location

India

Challenges

• Data missing for 50% projects
• Data mismatches
• One-third of the real estate project records could not be used for modeling due to missing price or inventory data
• Project amenities specified using free text; same amenity could have multiple descriptions

Solution

• Software developed on R Shiny platform
• An NLP technique, Word2Vec was used for specification and amenity data
• Clusters of micro markets were formed by hierarchical clustering method
• Algorithm for forecasting velocity of sale was developed
• Link with the database for comparison of any new project

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Portfolio inputs

• Maximize expected return by using optimization heuristics to solve portfolio optimization problems with different measures of risk (Variance, Semi-variance & VaR)  and multiple real-world constraints like Budget, Holding size for each asset, Trade limits for each asset, Cardinality, Round lots, Short Sales, Turnover, Beta etc.

Solution

• The portfolio optimization tool is a state-of-the-art tool for portfolio managers to arrive at the most profitable portfolio while meeting all business constraints. It uses the following heuristics to arrive at the optimal profitable portfolio

• Simulated Annealing
• Tabu Search
• Genetic Algorithm

Project information

Techniques

Mathematical Optimization

Client

Leading Broadcasters in India

Industries

Financial Services

Location

India

Portfolio inputs

• Asset Index
• Score
• Sector & Country Index
• Price
• Initial Investments (Units)
• Trade Limits
• Min-trade
• Min-holding
• Min-holding
• Benchmark-weight
• Beta

Other inputs

• Limits for country indices
• Limit for sector indices
• Covariance matrix and VAR of assets
• Returns of the assets in various historical periods
• Others

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!

Issues & Objectives

• Creating advertising proposals is a vital aspect of a broadcaster’s operations, as it centres around persuading advertisers to commit to investing in advertising slots or campaigns on the broadcaster’s platforms. These proposals encompass different rates for different advertisers based on the frequency and quantum of ad bookings
• Ad Revenue Optimiser (ARO) is a web application to automate and optimise advertisement inventory planning for the client resulting in additional revenue gain. The solution focuses on creation of proposals, planning of ad inventory and post evaluation

Project information

Techniques

Mathematical Optimization

Client

One of Leading Broadcasters in Asia

Industries

Media

Location

India

Solution

Designed and developed a software serve clients evolving requirements like

• State of the art inventory visualisation
• Advertising on digital and mobile platforms
• Interactive content
• Comprehensive sales management

Benefits

• Improved inventory pricing
• Improved allocation of inventory
• Improved pipeline visibility
• Sales executives performance tracking
• Negotiations history tracking for future reference
• Improved handling of Make goods
• Streamlined, accurate and faster billing

What we build, how it performs – Explore our work!

Our Case Studies

Would love to hear your thoughts!