Every time I raise an analytics use case, IT responds with a 5 months project and 500K costs... All I need a sandbox! (CMO)
Type: Product
A product that leverages machine learning techniques and helps business owners or production engineers with industry and market expertise to experiment with their data and get immediate, intuitive and actionable insight (e.g. production optimization, cross-sell, purchase propensity, churn propensity, etc.).
Service Mgmt. & Support
Hotline (English):
- Business hours (GMT +03:00, negotiable).
- Remote support provided via ticketing system.
- Email groups are available to post questions (counts as a ticket).
- SLA defined for critical/major/minor issues.
- No additional difference for bronze/silver/gold service levels.
Competitor Product
- SAS Enterprise Miner
- RapidMiner
- DataRobot
- Alteriycs
- Power BI ML
Use Cases & Pain Points Addressed
This tool solves the following pain points, or greatly reduces their impact:
Data analytics projects are expensive, and results are not known.
With this tool...:
Every business user can perform machine-learning analyses on a prepared sample of tabular data of any kind (e.g. about customer data, products, company, sensors...).
The complexity of algorithms and scores of Python and R codes have been translated into an easy-to-use application with comprehensive dashboards. No need for sophisticated theoretical background.
With a few clicks, and within a few minutes, data can be uploaded and processed to extract future predictions & newly found business parameter values.
Churn, Cross-sell, Fraud and many more data-driven predictions for relevant business questions can be addressed without training or IT-induced latency.
Examples of business scenarios and achievements (non-exhaustive):
Data analytics projects are expensive, and results are not known.
With this tool...:
- Quick trials are possible, at no cost.
- Results can be seen, before the project starts! (you get what you see)
- There is neither need for IT support to integrate the results, nor for a BI Tool to visualize the results. Output shows immediately to the user, with clear, easily understandable parameters and intuitive charts.
Every business user can perform machine-learning analyses on a prepared sample of tabular data of any kind (e.g. about customer data, products, company, sensors...).
The complexity of algorithms and scores of Python and R codes have been translated into an easy-to-use application with comprehensive dashboards. No need for sophisticated theoretical background.
With a few clicks, and within a few minutes, data can be uploaded and processed to extract future predictions & newly found business parameter values.
Churn, Cross-sell, Fraud and many more data-driven predictions for relevant business questions can be addressed without training or IT-induced latency.
Examples of business scenarios and achievements (non-exhaustive):
- In a Fortune 150 Payment Services company, revenue loss from churn was reduced by over 70%. It clearly identified churn characteristics and allowed to take preventive action.
- A retail chain gets real-time (predictive) information about the next best product recommendation, at the cash-it time, through API integration.
- Government's effort to fight business fraud and smuggling was predicted with over ten-fold accuracy compared to previous methods.
- For an OTA, target groups for a specific e-mail campaign could be predicted & segmented, leading to click rates 17% higher vs. a control group...
Key Features & Differentiators
The following features and advantages should be noted:
An intuitive & non-technical user interface, helps focusing on business value instead of technical details.
Simplicity: the platform was designed by predictive modelers, so the modelling flow is smooth.
None of them are as interactive and transparent:
No IT footprint: the platform can be used over the cloud and web, unless required by the policies of the enterprise.
Costs/Expenses: competitors are either more expensive and/or have fewer capabilities.
Performance benchmark:
An intuitive & non-technical user interface, helps focusing on business value instead of technical details.
Simplicity: the platform was designed by predictive modelers, so the modelling flow is smooth.
- It has a unique responsive app interface: one can log in and build models on a smart phone or tablet.
- After uploading data, a fully transparent machine-learning model can be built in 3-clicks.
- Advanced users can then move on to more advanced features and settings in case of need.
None of them are as interactive and transparent:
- None support playing with parameters in what-if scenarios... literally while sitting in a CXO meeting.
- Others do not provide trials and do not open the platform to public for unsupervised use.
- More than ""try & buy"": the tool's web-platform is live, ready to sign-up and use. Only Dataiku provides such a service (though via download to local pc).
No IT footprint: the platform can be used over the cloud and web, unless required by the policies of the enterprise.
Costs/Expenses: competitors are either more expensive and/or have fewer capabilities.
Performance benchmark:
- In one large customer, prediction of customer churn has an error variance of only 1%.
- 10 minutes are enough to learn from 1 Mio records (data points).
- For smaller data sets with less than 100K records, it runs in seconds.
