Chatterbox Labs’ software products are built to work immediately within a standard Enterprise software environment. There is no requirement for bespoke hardware or software, all products can be deployed on commodity server hardware and run on the standard JVM.
It takes minutes to deploy our software inside your firewall. Updates can be pulled via secure online repositories in line with your business objectives.
Patented Synthetic Data Generator
The Synthetic Data Generator (SDG) was built to:
- Alleviate the burden of preparing machine learning ready data using Patented Reinforcement Learning methods
- Delineate IP for data reusability & ownership using Sequential Regression with Multiple Imputation methods
- Create vast datasets for use in regulated environments using the NLP Text Generator
The SDG addresses 1) poor data, 2) small data and 3) no data.
SDG enables real-world Enterprise AI projects where data is the constraint.
Patented Cognitive Engine (AutoML)
Chatterbox Labs’ Auto ML for classification, regression and time series forecasting was built for Subject Matter Experts, Data Architects, Data Analysts and Software Architects with no formal machine learning background.
In order to solve real world problems across varied data types and industries, a plethora of machine learning methods are automated within the engine: Neural Network, Support Vector Machine, Naïve Bayes, Random Forest, Latent Dirichlet Allocation, LASSO, Edge, Ridge, Elastic Net, Exponential Smoothing & ARIMA.
The compelling attribute of our Auto ML is the subject matter expert, architect or analyst does not need to interact with any underlying machine learning. Operational execution is simple to absorb; a 3-hour workshop will provide the necessary training users require to deliver real-world AI outcomes.
Chatterbox Labs’ Explainable AI was built to alleviate an Enterprise’s frustration with black box outcomes.
Our approach is not to try to solve the technical intricacies of the underlying machine learning algorithm. Whilst we use ground breaking methods including Ablation, Distance Measures and Influence Functions (akin to Google Brain research), our focus is on delivering interpretable outcomes that business leaders require in order to justify the next best action whether that be fully automated or via manual intervention.
Explainable AI comes into its own within regulated and complex environments such as Healthcare, Financial Services and Pharmaceuticals. Identifying salient features within complex data points arms a human with the justifications needed to determine the next best action in their business process.