Making a Difference

We adopt a bottom-up approach starting with strategy support through discovery and analysis of existing systems and processes and their limitations maximizing the leverage of knowledge and results in higher value generation!

PROBLEMS ADDRESSED

Case Studies

PROPENSITY MODELLING & RECOMMENDATION SYSTEM

REQUIREMENT

Deliver a personalised recommendation experience to customers by displaying relevant products in insurance, healthcare and retail based on customer's purchase behaviour, past sales data, personal interests, gender, demographics and likelihood to buy based on intent signals.

SOLUTION APPROACH

Designed propensity models and recommendations leveraging ML tools and techniques.

HIGHLIGHTS/TECH COMPONENTS

Leveraging techniques like Machine learning, Random Forest, XGBoost collaborative filtering, content based filtering, hybrid filtering, deep learning.

BENEFITS OBSERVED

Improved sales execution to enable cross-sell and up-sell, effective use of resources, improved campaign effectiveness, reduced costs and provided intelligent pricing.

PREVENTIVE MAINTENANCE

REQUIREMENT

Develop an automated process solution that continuously monitors sensor data to predict upcoming downtime/faults and raise tickets ahead of time.

SOLUTION APPROACH

Developed classification and regression models using data from IOT sensors.

HIGHLIGHTS/TECH COMPONENTS

Leveraging sensors and machine learning techniques like classification and regression.

BENEFITS OBSERVED

Minimize downtimes, increased customer satisfaction, reduced losses, maximized asset lifetime.

VIRTUAL SHOPPING ASSISTANT

REQUIREMENT

Develop a virtual shopping assistant to help shoppers find the best products in the easiest way across multiple domains like retail, eCommerce and healthcare.

SOLUTION APPROACH

Conversational AI & Visual Search to help users intuitively find what they're searching for.

HIGHLIGHTS/TECH COMPONENTS

Deep learning techniques like Encoder-Decoder models, RNNs & deep Seq-2-Seq models to create conversational assistants. Used deep CNN to create image tagging and openCV\CNN classification to match similar products to implement visual search.

BENEFITS OBSERVED

Enhanced customer shopping experience by providing customer centric search. Could run personalized loyalty programs that resulted in improved sales and customer loyalty.

RISK ANALYTICS

REQUIREMENT

Client mandated a solution that can automate the lending process, automate credit checks, design alerts to monitor anomalies and outliers, identify high risk customers, screen for risky deals, simulate portfolios and evaluate potential impact of possible trades.

SOLUTION APPROACH

Developed classification models using Machine-learning techniques, such as deep learning, random forest, and XGBoost.

HIGHLIGHTS/TECH COMPONENTS

Machine-learning techniques like deep learning, random forest, and XGBoost.

BENEFITS OBSERVED

Reduces risk of loss, higher interest incomes, lower sales and operating costs, reduced costs associated with risk mitigation, improved capital efficiency.