Problem statement: Our task involved creating a structured dataset that contains 6 million news articles and can analyze the texts in seconds. Dataset: The dataset is set of real-time trackers which ‘intelligently’ filter relevant articles from US, German and Indian newspapers and add them to the dataset. Our approach: Data gathering: Our algorithm automatically scanned newspaper websites and stored the article...
Problem statement: Our client wanted to extract insights about competitors and internal processes based on external communication with potential or existing clients. Dataset: The dataset is set of real-time trackers which ‘intelligently’ filter relevant articles from US, German and Indian newspapers and add them to the dataset. Key value addition: Proprietary textual classifiers (enhanced Bayes classifiers) developed in-house using SCALA/MongoDB were...
Problem statement: We were commissioned by an e-commerce platform to conduct an analysis on client reviews with regard to their satisfaction. Dataset: Data-collection: Our algorithms automatically collected client reviews from webpages and stored them. Data structuring: In order to analyze the sentiment of a reviewer, the stored text was structured according to different word and syntax characteristics Machine learning: Once...
Problem statement: For a retail firm we were commissioned to analyze the sales growth evolution as well as creating forecasts. Moreover the client asked us to analyze the composition or regional sales and the importance of regional shops. Dataset: Re-structure data: In order to make the analysis possible we re-structured the data from the internal ERP system. Identifying correlations: We...
Problem statement: Our client wanted to understand baskets of products customers tend to buy together, depending on their age, gender, purchasing power along with other characteristics such as weather, time of the day, month or year. Dataset: The dataset consisted of approximate 5 million transactions, 600,000 receipts and 10,000 customer card holders. Our approach: Advanced mathematical models based on Class Association...
Customer Challenges: Sluggish user interface, spaghetti code. Existing team was non-existent CoreView needed to build an engineering team to work on multiple fronts. Maintain existing product, add new features, and modernize Slew of new features from PM, meant new challenges, new technologies to understand and add features in timely fashion Make new features agnostic to the underlying automation platforms...
Customer Challenges: Platform independent solution for Windows OS specific file formats Stringent performance and scalability requirements Automation and integration with existing product with at par functionality Highly complex and interdependent internal structure of the source( PST, MSG ) Result: 25x improvements in reading data from PST and MSG files Libraries built in a way to scale up and scale out...
Customer Challenges: Ensure data integrity while satisfying business continuity requirements Support existing customers as well as bring the new development at par with the existing functionality Become source of truth for all applications connecting to system Near real-time add / update of customer data Hugh amount of data: 4 million organizations, 10 million organization names, 7 million consumers, 320...
Customer Challenges: Built services based on AWS serverless architecture Service architecture based on Lambda functions UI technology upgraded to React JS Built reusable UI components like customizable grid and calendar Integration of third party services like reputation management Result: Improved third party integrations Better use of AWS services CoreView Solution: Built services based on AWS serverless architecture Service architecture based on...
Customer Challenges: Needed replacement for a very expensive software subscription service Did not have the technical competence to build such a solution Needed someone to Build-Operate & Transfer the technology, not team Result: Instant cost savings by one M $ every year. In premise deployment - Increased Security of customer data 99.9% uptime maintained for over three years by now ...
