Résumé — Anusha Seshadri

AI / Data Scientist leading the AI and data science function at Digestive Disease Care, applying deep learning, generative AI, and applied statistics to healthcare operations, claims, and interoperability problems. M.S. in Data Science from Rochester Institute of Technology. Holds an Indian patent for a novel AI-driven clinical risk prediction framework applying optimized membership functions and defuzzification to multi-dimensional biological attributes. Strong quantitative and programming foundations in Python, PyTorch, and CUDA, with experience analyzing large-scale imaging-adjacent and sensor datasets in multidisciplinary research settings.

Experience

AI EngineerJun 2026 — Present

Digestive Disease Care — New York, NY

Leading an AI and data science initiative focused on claims and revenue cycle operations, including a denial-prediction model, automated insurance verification at intake, and payer performance analytics, in active discovery with billing and clinical teams to map current systems and historical claims data.
Scoping an interoperability layer to standardize payer data exchange, evaluating HL7 FHIR-based integration with the practice's Epic environment to support real-time eligibility checks and cost estimation.
Designing an OCR-based insurance card intake pipeline paired with DDC's historical billing data to generate pre-visit out-of-pocket cost estimates for patients.
Applying prior AI/ML research experience (deep learning, NLP, generative AI) to healthcare-specific problems, including claims denial patterns, patient billing communication, and operational analytics.
One of two founding members of DDC's AI/Data Science function, positioned to lead the team as it scopes and hires additional engineers to build out the roadmap.

AI / ML Research ScientistJan 2025 — May 2026

Rochester Institute of Technology (NGAFID)

Developed and applied advanced AI and deep learning methods to research questions in large-scale sensor and time-series data analysis, building and optimizing computational approaches for anomaly detection and pattern recognition on real-world operational datasets from the NGAFID aviation database.
Designed, built, and trained deep learning models using PyTorch and CUDA on NVIDIA GPU, training 10,000+ neural network architectures in parallel via neuroevolutionary search (EXAMM), applying rigorous quantitative analysis and statistical validation to extract meaningful patterns from complex, high-dimensional sensor data.

Graduate AssistantAug 2025 — May 2026

Center for Leadership and Civic Engagement, RIT

Applied generative AI and foundation model approaches (Anthropic Claude API) to develop new computational methods for workflow automation, contributing reusable AI patterns and collaborating closely with cross-functional teams to integrate AI capabilities into research and operational workflows.

Data Science InternDec 2023 — Feb 2024

TATA Elxsi

Developed and applied machine learning methods to large-scale sensor and behavioral datasets, building computational approaches for classification, anomaly detection, and pattern recognition that improved system efficiency by 15% and reduced errors by 20% in a collaborative, multidisciplinary engineering environment.

Applied Data Science InternMay 2023 — Jul 2023

Smartbridge · Google Developers

Designed and deployed deep learning-based prediction pipelines from data ingestion through model serving, applying neural network architectures and generative AI-adjacent modeling techniques in a collaborative research-to-production workflow.

Projects & research

shadow360: Healthcare Career-Exploration PlatformDigestive Disease Care

Designed and built a server-authoritative assessment and eligibility engine that scores exams, enforces timing, and gates shadowing access based on rule-driven, per-student logic, replacing a client-side prototype with exploitable scoring and access gaps.
Built a resumable assessment pipeline that draws and shuffles per-attempt question snapshots from a question bank and issues immutable certification records that persist across retakes.
Implemented data protections for a minor-heavy user base, including JWT auth with concurrency guards, per-account and per-IP rate limiting, and audit logging, aligned to NY Child Data Protection Act and SHIELD Act requirements.

YogiSync: Real-Time Computer Vision and Pose Recognition System

Built a real-time computer vision pipeline using OpenCV and MediaPipe for live skeletal keypoint extraction, pose landmark detection, and multi-class pose classification from continuous video streams, training a custom Random Forest model on a self-collected diverse imaging dataset with rigorous preprocessing and validation.
Integrated generative AI (Google Gemini LLM) for personalized coaching and multimodal interaction, applying deep learning and image analysis techniques to extract meaningful patterns from live visual data streams in real time.

Predictive Anomaly Detection in NGAFID Aviation Data — Deep Learning Research

Architected a 2-stage deep learning research pipeline combining One-Class SVM and neuroevolutionary RNNs for anomaly detection on large-scale real-world time-series sensor data, applying advanced quantitative analysis, statistical modeling, and explainable AI methods to identify complex temporal patterns.

MendPath: Generative AI and Multimodal Health Application

Built a generative AI-powered health application integrating voice AI, multimodal LLM interaction, and RAG-based personalized recommendation pipelines — applying foundation model approaches and generative AI to biological and behavioral health data in a research-to-product development workflow.

Make Ideas into Motion — GAN-Based Generative Imaging Research

Designed and implemented a generative adversarial network (GAN) pipeline for converting natural language descriptions into motion imagery, applying deep learning, data augmentation, and diffusion-model-adjacent generative techniques on cognitive and neurological imaging-adjacent datasets — published in a peer-reviewed journal.

Patent & publications

Patent (Indian Patent Office) App. No. 202441004228Filed Jan 2024 · Published · Under Examination

Hybrid Fuzzy-Logic Clinical Risk Prediction Framework for Aging-Relevant Disease: Developed a novel AI system for cardiac risk prediction in clinical populations using Particle Swarm Optimization (PSO) to tune fuzzy membership functions over multi-dimensional biological attributes, with centroid-based defuzzification for interpretable, quantitative risk scoring. Applied advanced feature selection pipelines and statistical modeling on real clinical datasets to identify biological risk patterns, producing a scalable and interpretable clinical risk scoring framework.

"Make Ideas into Motion using AI"Journal of Software Engineering and Simulation, Vol. 9, Issue 4 (2023)

Applied NLP, deep learning, data augmentation, and generative adversarial networks (GANs) to convert natural language descriptions into motion imagery, contributing to generative AI research for cognitive and neurological applications.

Technical skills

Healthcare & Interoperability: HL7 FHIR, Epic, EHR/Payer Integration, Insurance Claims Analytics, Denial Prediction, Revenue Cycle Analytics, OCR Pipelines, Healthcare Data Compliance (HIPAA, SHIELD Act)

Deep Learning & CV: PyTorch, TensorFlow, Keras, CNNs, RNNs, Transformers, Computer Vision, Image Recognition, Segmentation, Pose Estimation, Skeletal Keypoint Extraction, Object Detection, OpenCV, MediaPipe

Generative AI: GANs, Diffusion Models (foundations), Foundation Models, Multimodal Models, LLMs, Anthropic Claude API, OpenAI, LangChain, LangGraph, RAG Pipelines, Generative AI Pipelines

Quantitative & ML: Statistical Modeling, Feature Engineering, Anomaly Detection, Causal Inference, Probabilistic Modeling, Scikit-learn, Neuroevolution, Fuzzy Logic, PSO Optimization, Explainable AI

Infrastructure: CUDA, NVIDIA GPU, Docker, GitHub Actions, CI/CD, AWS (S3, EC2, Lambda), GCP, Apache Spark, Weights and Biases, Experiment Tracking

Languages: Python, SQL, Java, TypeScript, JavaScript, C++, Scala

Research: Peer-Reviewed Publication, Patent Author, Multidisciplinary Collaboration, Quantitative Data Analysis, Large-Scale Dataset Processing, Reproducible Research

Anusha Seshadri

Experience

Projects & research

Patent & publications

Technical skills

Education