Keep in touch with meI'm using Intch to connect with new people. Use this link to open chat with me via Intch app
Work Background
IBMData Scientist
Aug. 2023New York, New York, United StatesI’m a Data Scientist working in client-facing environment where analytics, applied AI, and system design overlap.
At IBM, I design and implement Python-based analytics, automated workflows, and cloud-native data pipelines that support Medicaid fraud-risk detection for state agencies operating at scale. Most of my work involves taking large, messy healthcare data and turning it into usable signals by identifying irregular billing patterns, surfacing financial anomalies, and improving scalability and operational efficiency. I’ve also worked on modernizing legacy analytics platforms by integrating Generative AI, LLM-based workflows, prompt optimization, and intelligent automation into existing systems.
Additionally, a meaningful portion of my role functions in practice as solutions-engineering work. I work closely with program managers, policy leaders, and technical teams to understand regulatory requirements, data constraints, and business goals, then translate those inputs into end-to-end analytical and AI solution designs. This includes defining system logic, data flows, validation frameworks, and API-based integration points, as well as supporting client-facing conversations by explaining AI, LLM, and analytics capabilities in a way that’s clear and grounded for non-technical stakeholders.
Furthermore, I’m proficient in AWS and have hands-on experience deploying analytics and AI workflows within cloud architectures, with additional exposure to Azure and its AI ecosystem. I regularly contribute to solution-level discussions around data ingestion, processing, model execution, automation, and output delivery to support scalable, enterprise-ready platforms.
I’ve supported client-facing POCs, analytics engagements, and solution demos across multiple state healthcare programs, and regularly present analytical and AI concepts to both technical and nontechnical audiences. I’ve also designed and delivered enterprise product trainings to support adoption and long-term use.
IBMAssociate Data Scientist
Aug. 2022New York, New York, United StatesI collaborated cross-functionally with engineering, analytics, and product teams to support the development and ongoing enhancement of an enterprise business automation platform. My work involved leveraging SQL for large-scale data querying, validation, and transformation across complex datasets, ensuring data integrity and consistency within automated analytics and decision-support workflows.
I stayed current with advancements in data science, Generative AI, and intelligent automation, incorporating emerging best practices to continuously enhance analytical capabilities and platform performance. This included contributing to the evolution of analytics workflows designed to support scalable, repeatable insights across business and regulatory use cases.
In addition, I supported code review, testing, and refinement efforts for the Intelligent Automation Platform, working closely with technical teams to improve reliability, maintainability, and scalability. I also contributed to the development of Business Question Prompt Optimization and automated prompt generation processes, including LLM-enabled workflows, helping standardize how business questions are translated into analytics and automation outputs.
Together, this work supported more consistent, AI-driven analytics delivery and strengthened the platform’s ability to power enterprise-scale business automation and decision support.
TakedaData Engineering and Artificial Intelligence Intern
Jun. 2021 - Aug. 2021Boston, Massachusetts, United StatesI designed a data pipeline to ingest large amounts of data quickly and provide exploratory data analysis for each patient and to get the data ready to be put into a neural network to predict target symptoms. I designed this data pipeline using PySpark and partioned the data, which was 1.8 terabytes of data, using a Resilient Distributed Dataset to drastically decrease the run time of the pipeline. The pipeline that I designed decreased the processing time from originally taking 4 to 5 hours per patient to a process that took under 20 minutes per patient. I applied my knowledge of set theory, indexing, and data structures to create this pipeline. The pipeline is now automated to be used once a month to continually generate summary statistics once a month.
After the pipeline, I also created a classification model that would predict target symptoms that would then be used to see when the patient should be warned to take their medication to prevent the symptoms. I used a neural network to create the model.
Cornell UniversityUndergraduate Research Assistant
Aug. 2019 - May. 2022Tompkins County, New York, United StatesI worked under Dr. Cheng Zhang in the Computing and Information Sciences Department at Cornell University. I developed a sensor that predicts hand-to-face touching using machine learning to prevent excessive face touching. In order to create the sensor, I needed to analyze large amounts of data using Python with the NumPy, Pandas, and Sci-kit Learn packages. Additionally, I used neural networks using the Keras package to create the model that would prevent excessive face-touching by using machine learning to predict and then warn the user that they were about to touch their face.
I have also worked on a second 3D-body modeling project. This project's goal was to predict full body movement from wrist sensors. For this project, I first helped to collect and transform raw data into data that was used as training or cross-validation, or testing data using NumPy and Pandas. I then helped to create the model that would predict the full body movement using TensorFlow.
Virginia Commonwealth University - College of EngineeringHigh School Research Assistant
May. 2017 - May. 2018Richmond, Virginia AreaI worked in Natural Language Processing (NLP) Lab at Virginia Commonwealth University's Computer Science Department. I developed a package using Perl and Javascript to cluster and visualize alike terms by using unsupervised machine learning to discover relatedness of drugs and diseases using Literature Based Discovery techniques.
Virginia Commonwealth University - College of EngineeringHigh School Research Assistant
May. 2017 - May. 2018Greater Richmond RegionI worked in Natural Language Processing (NLP) Lab at Virginia Commonwealth University's Computer Science Department. I developed a package using Perl and Javascript to cluster and visualize alike terms by using unsupervised machine learning to discover relatedness of drugs and diseases using Literature Based Discovery techniques.
Intch is a Professional Networking App for the Future of WorkSee more people like Samhita on Intch
300k+ people
130+ countries
AI matching
Get Intch if you'd like to
499364 people
39
CEO @ Papillon
15
Production Assistant
16
Program Manager @ DISH Network
221534 people
39
CEO @ Papillon
15
Consultant for construction project management and construction company turnaround. @ Self Employed
15
Entrepreneur @ Becoming Your Own Boss