Apurba leverages the power of Natural Language Processing (NLP) to understand textual and tabular data to derive actionable intelligence.
Apurba has developed a framework that we call Big-Data-in-a-Box which can be adapted to various applications and platforms with relative ease.
Apurba exploits the power of Machine Learning to model your complex data, solve your big puzzles and answer your tough questions.
We built an insightful, patient-specific narrative based on contextualized, domain-specific models for different disease states based on clinical texts about that disease. This model can be used to generate customized patient summaries from the patient’s electronic health records, dynamically presenting the most pertinent information for any medical situation, ranging from regular appointments to ER situations.
In another case, we built a solution that can process unstructured, free-format text stored within Patient Notes and aggregate that with the structured data located within tables and charts. Apurba is spearheading the new wave of innovative technologies to facilitate the transition from paper-based records to electronic records, specifically for healthcare providers. Apurba’s NLP solutions can fundamentally change the way the industry interacts with patient records and improve clinical outcomes.
We have built a ‘Smart Room’ to collect and analyze multi-modal doctor-patient interaction data within a Big-Data analytics system. In modern healthcare systems, data is disjointed and disparate, making key factors in caring for the patient difficult to measure and correlate, resulting in unnecessary procedures, additional hospital visits, and extra costs for both patients and the healthcare organizations that serve them.
This will solve this problem by building a system that can collect and simultaneously process multi-modal healthcare data, such as audio (e.g. speech, tone, sentiment), movement (e.g. gait), video (e.g. face recognition, pain estimation, sentiment analysis), and derived signals (e.g. speech-to-text natural language analysis), within the same capture and processing framework. This will revolutionize the way healthcare information is analyzed, opening up a new set of tools to effectively comprehend complex, interdependent medical data to derive actionable intelligence.
Apurba just recently successfully completed Phase I of a Big Data analytics contract to model heat characteristics of engine exhaust temperature based on the client’s internal testing dataset. This solution creates a data-driven thermal model of an engine based on nothing but a set of previously run test runs using machine learning and Big Data analytics. Once the model is built, the model can, with 90% accuracy, predict the exhaust temperate of the engine given a random combination of random variables such as ignition time, fuel ratios, barometric pressure, humidity, starting temperature, etc.
Another example is a Big-Data Analytics Platform that leverages machine-learning algorithms to identify the most clinically effective specialty medications at an optimal cost for individual patients, which can then be extended to execute cost-benefit analysis on the behalf of entire medical institutions. This platform is ushering in a new era of precision medicine, providing actionable intelligence that enables healthcare providers the ability to prescribe exactly the right drug, for the right patient, all at an optimized value.
Years of Experience