' '




Customer Sentiment Analysis

The solution pulls in data from various social media data sources such as Skytrax and provides insights into the customer sentiments that can help shape the customer service strategy for the airlines. With airlines trying harder than ever to retain loyal customers, customers relying heavily on social media and internet in their decisions, and customer demanding excellent service for the price they pay, it is imperative for any airlines company to track the customer sentiments about its brand.

Our solution first classifies each comment under various topics such ascatering, booking, comfort, staff, entertainment, etc. using a Naive bayesian classifier. Within each topic bucket, a sentiment analysis engine leverages a language semantic disctionary as well as a domain and topic specific ontology to gauge the sentiments associated with the comments. The solution can pull data for the airlines as well as its competitors to gauge relative sentiments associated with certain topics and this provides significant inputs for driving competitive strategy.


Ancillary Revenue Analytics

Ancillary revenue is a significant profitability driver, particularly for a low cost carrier. The solution analyses historical transactions of individual customers and does cluster analysis based on buying patterns and demographic information to understand the propensity of individual customers as well as customers within an action cluster for buying a certain ancillary item. The propensity scores facilitate personalized target marketing recommendations for ancillary products that increases the overall conversion. A feedback look updates the propensity scores based on every deviation between observed behaviour and expected behaviour of any customer. Thus the recommendations engine remains dynamic and the machine learning approach facilitates increased accuracy in propensity modelling with increase in usage.


Revenue Target Monitoring

The solution provides the airlines' performance monitoring algorithm for any station, route, region or the entire network. It integrates data from the Reservations, Forecasting and Budgeting Systems, IATA PaxIS, providing a holistic view of the business to the management. The solution included a predictive component that forecasted revenue for a defined future time period taking into account observed seasonality and demand variability across clusters over time. Past performance data and future revenue predictions helped in reactive as well as proactive course corrections.


Analytics Around Flight Disruptions

The solution tracks the impact of minor disruptions on various customers, focusing on the impact on frequent flyer members at various levels.The disruptions can be due to different reasons such as Late Check in, Airport Disruption, Weather disruptions, Equipment Loading etc. The reasons can be analysed. In addition, the solution tracks the impact of disruptions on future buying behaviour of customers and provides insights that can help the airlines retain valuable customers who are showing a propensity to churn, post a disruption.


Distribution Channel Analytics

This solution represents the revenue and cost of distribution of tickets via different indirect and direct channels, such as GDS, OTA, Call Center, ATO/CTO, Mobile/Website etc. As certain channels are more profitable than the others, the analysis can help airlines identify a set of passengers and strategize in order to move them to lower cost channels.


Airport Operational Analytics

The solution provides insights on the operations performance analytics for a major airport, and highlights the overall operational efficiency at the various passenger touchpoints - Check-in, Arrival Immigration, Departure Immigration and Baggage Belt. The solution compares the actual processing time with set benchmarks and provides the ability to analyse cases where the processing time exceeded the set threshold. The solution can also highlight the result of customer experiences as captured by structured/semi-structured surveys.


Route Profitability Analysis

The solution provides analysis on the route profitability that includes the identification of the most profitable/non-profitable routes, and also compare the profitability of a route/leg vis-à-vis the revenue generated by the same. The solution helps the airlines in identifying high revenue opportunities through a detailed analysis of its forward booking window in culmination with historic trends observed.


Optimal Allocation Plan of Vehicles

The solution recommends an optimal allocation plan of vehicle type, having a relevant capacity and cost of operation, to the operating routes of a Vehicle Company possessing a fleet of different vehicles types in their inventory with the objective of minimizing the expected total cost of operation. This is done by anticipating the demand pattern on the operating routes and deciding which vehicle should be allocated to which route in order to minimize the total cost of operation. The optimization algorithm uses mixed integer programming.


Optimal Route Network Determination

The solution determines the feasible aircraft routes network such that each flight leg (an Origin Destination Pair) is covered exactly once by an aircraft, respecting constraints such as the maintenance requirements for an aircraft type, the number of aircraft available in the inventory, the schedule of flights, etc. The optimized solution avoids short connections to the extent possible, and maximizing the number of connections that are attractive to the passengers. The model is solved for a planning horizon using integrated heuristics and mixed integer programming approach.


Ticket Pricing Analytics

Dynamic pricing solution developed for a US based full service carrier, using statistical pricing models. Model takes into account historical booking patterns across various sectors, time, day of week, festivals, etc to forecast the expected ticket sales in the forward booking window, which in turn feeds into the pricing engine to recommend optimal prices for the airlines.


Minimum Cost Vehicle Maintainence Plan

The solution recommends a minimized cost solution for vehicle maintenance planning where the aim is to find the optimum balance between the costs and benefits of maintenance, while taking all kinds of maintenance policy constraints into account. The solution considers the maintenance cost of a vehicle of a specific age and takes into account factors such as failure, ageing, inspection repair, replacement, system life, etc. and finds the optimal maintenance cost limit that minimizes the expected total cost of maintaining and replacing a vehicle over a fixed planning horizon. The analysis involves solving a stochastic optimization.


Fuel Over Destination Analysis

Fuel Analytics Solution optimizes the amount of fuel carried by each flight. Our solution consolidates information from multiple data sources such as Flight Plan, Flight Operations, MRO and then predicts the optimal fuel required for each and every flight. Parameters analysed include Fuel Utilization Trend Analysis, Fuel Burn, Factor for head-wind, weather patterns for in-flight route changes, Route optimization, Load Planning, Burn Variance and Maintenance Information. Optimal fuel capacity is recommended for a flight based on advanced optimization models with applied heuristics.


Device & Application Log Analysis

The solution analyses streaming data for different devices to identify usage patterns across a time period. Detailed analysis can be done to determine device performance and identify cases that have the highest or lowest response times. The solution used a kafka integration to facilitate the real time data ingestion.


Eco-Tasks Analytics

The solution developed for a major full service carrier analyses the performance of various facilities for performing eco-tasks. The solution provides insights on how their different facilities are performing in various regions using key metrics such as number of tasks completed on time, number of tasks overdue, number of tasks missed, etc. A logistic regression based model predicts the probability of a certain job type being completed in a given region at a given time for the relevant employee pool. This helps correct allocation of tasks and minimizes overall delays in task execution.