Project Team

Kubrick unlock actionable insights and trends by uniting fashion sales and social media fashion sentiment

• Merge of multiple social media streams and internal sales data to show accurate performance of sales vs public fashion sentiment

• Consolidation of all data sources into easily queryable data lake

• Production of simple to use front-end platform for analysis (website front-end)

• Concept now being productionised to better predict stock levels and reduce waste

Kubrick enable InsureTech Company to predict credit score and delinquency rate with up to 85% accuracy.

• Production of predictive engine to return accurate delinquency and credit scores

• Identification of key credit score change parameters

• Consolidation of all data sources into an easily queryable data lake

• Analysis of company bankruptcy trends identifying tell-tale signs of company failure.

Kubrick utilises Text Analysis to drive highly complex website structure

• Detailed interactive map of website including all documentation and interrelationships of information types

• PDF text parsing using Python

• Document clustering using un-supervised machine learning

• Document hierarchy using text analysis & Google Analytics

Kubrick drive insurance risk assessment by harnessing the power of external data

• Production of cyber risk assessment engine.

• Scraping & engineering of numerous external data sources to provide improved insights into insurance risks.

• Social media scraping engine & relationship mapping to detect fraud

• Production of predictive model to predict cyber data breaches

Kubrick derive customer honesty through deep analysis of customer behaviour

• Data quality grading based on analysis of over 60 million transaction records

• Ability to produce personalised advertising campaigns based on granular segmentation.

Company understands employee sentiment better through analysis of open source company data

• Production of topic modelling engine to analyse employee reviews

• Scraping of numerous online data sources to produce a deep view of employees

• Production of flexible, interactive web-based dashboards for executive interrogation.

Data Driven Approach to handling bad publicity

• Automatic ingestion of multiple online data sources

• Sentiment analysis on free-text data

• Linking & network analysis between journalists & publishers

Natural Language Processing driven web content restructuring

• Automatic scraping and logging of all website locations & paths

• Visual mapping of all website interactions and locations # topic analysis & NLP to determine website re-structuring

• Google Analytics deepdive

Employee Churn analytics

• Scraping and aggregation of employee review websites

• Normalization & NLP of reviews

• Automated reporting and analytics of results

Trend analytics in retail correlated to social media

• Social media scraping & aggregation

• NLP and trend analysis on both internal and external data

• Automated reporting and analytics to drive marketing campaigns & seasonality understanding.