The Unseen AI: How Data Integrity Saves Lives in High-Risk Healthcare

A Profile on Prasanth Tirumalasetty: IFGICT Fellow Architecting Verifiable AI in MedTech

Prasanth Tirumalasetty, a seasoned Project Lead and Business Analyst III, operates at the critical intersection of high-risk medical device manufacturing, SAP enterprise systems, and advanced data science. His multi-domain expertise—spanning solutions architecture, data modeling, and regulatory compliance—positions him at the forefront of the industry’s most significant challenge. As a Researcher with a Master of Science in Data Science from the University of Michigan-Dearborn and a recognized expert in data lineage (ORCID: 0009-0001-1294-2375), he specializes in integrating the revolutionary speed of AI with the immovable mandate of FDA/ISO auditable compliance. This commitment to trustworthy data is underpinned by his intellectual property, including a Design Patent for the “Forma” Smart Recovery Mat and a filed utility patent application, “System and Method for Generating Privacy-Preserving Synthetic Health Data Using a Generative Adversarial Machine Learning Model,” which directly addresses core risks surrounding data utility and ethical usage.

Tirumalasetty argues that the conversation around medical Artificial Intelligence, while full of potential, often eclipses the single most critical challenge: technology must be auditable. In the high-stakes world of regulated medical device manufacturing, he insists that a “black box” is not merely a technical limitation; it is a profound regulatory and ethical liability. Tirumalasetty solves this paradox by architecting the Verifiable AI Ecosystem, where data integrity and compliance are automated from the foundation.

The Critical Challenge: Data Integrity and the Regulatory Black Box

Modern healthcare systems and manufacturing facilities are fundamentally reliant on complex, disparate enterprise software. Critical operational data—everything from raw material inventory receipt, instrument calibration, and component traceability to final quality sign-offs—is often fragmented and locked within these systems. The Enterprise Resource Planning (ERP) system, such as SAP, remains the single source of truth for manufacturing and logistics data.

This pervasive fragmentation creates systemic risk, compliance failure, and patient safety risks. The stakes are simply too high for manual checks and fragmented systems.

Solution 1: Auditable Foundations and Compliance by Design

The foundational step toward Verifiable AI is enforcing a “Compliance by Design” philosophy. This mandates that audit trails are built into the architecture, ensuring the digital infrastructure is inherently compliant and satisfies requirements like 21 CFR Part 11 for electronic records.

Drawing on his expertise across SAP and .NET environments, Tirumalasetty’s solution involves architecting secure, compliant data pipelines. Platforms like .NET Core are used to integrate seamlessly with core ERP systems, ensuring that every critical data point is captured in real-time, secured against tampering, and linked to an immutable record. This methodology is vital for capturing granular details and ensuring all serialization data is perfectly traceable. This continuous chain of custody transforms raw data into an auditable asset satisfying stringent FDA and ISO requirements.

Innovation Spotlight: The “Forma” Smart Recovery Mat Design Patent

Tirumalasetty’s commitment to compliance by design is exemplified in his work on the “Forma” Smart Recovery Mat, which is the subject of a Design Patent . This invention is a connected medical device ecosystem designed to solve compliance and low patient adherence in post-operative physical therapy (PT). The system functions as a closed-loop HIPAA-compliant device:

1. Unique Design and Function: The product features a unique, organic, teardrop shape and an integrated dome-shaped “Pod” Hub (the core of the design patent). This hardware uses piezoresistive pressure sensors and a haptic engine to guide the patient and instantly correct their form in real-time.
2. Verifiable Data Loop: The system utilizes a .NET Core backend (the “Forma” Cloud) to securely receive real-time movement data, which is then analyzed by a machine learning model for real-time pose validation.
3. Proactive Compliance: The system logs every time-stamped session into an immutable, auditable record. This verified, HIPAA-compliant data is then used by the clinician for Remote Therapeutic Monitoring (RTM) billing and clinical intervention, directly enhancing patient safety and generating new, reliable revenue streams for U.S. healthcare providers.

His broader technical frameworks in this domain include:

• “A Computer Vision and Machine Learning Framework for Automated Sterilization and Batch Validation”
• “A Generative Framework for Supply Chain Digital Twins”

Solution 2: Proactive Safety Through Predictive Verification

Leveraging his advanced expertise, Tirumalasetty builds predictive models designed to forecast potential compliance issues, predict manufacturing defects, or flag process anomalies before they manifest as a risk to the patient.

• His sole-authored manuscript on the “Deep Graph Learning for Autonomous Data Reconciliation (DGL-ADR)” framework is his core innovation in predictive data governance. This system is designed to forecast reconciliation errors across complex systems like SAP and Salesforce, acting as an always-on internal auditor.
• His innovation in Privacy-Preserving Synthetic Health Data is critical for training powerful models without compromising patient information.

This represents the fundamental shift from a reactive Quality Control (QC) mindset to a preventative Quality Assurance (QA) paradigm.

Conclusion: Global Standards and the Future of Trust

Tirumalasetty believes the future of medical technology relies on experts who can successfully bridge three distinct domains: cutting-edge academia, stringent regulatory compliance, and robust enterprise software engineering. His career demonstrates not only the ability to execute within these domains but also the sustained, national, and international acclaim that places him among the small percentage who have risen to the very top of his highly specialized field.

Sustained Acclaim: Recognition as a Global Authority

Tirumalasetty’s authority is consistently validated through multiple, independent external endorsements.

• Expert Judging and Keynote Speaking: Tirumalasetty has been recognized as an authority in innovation and technology evaluation. His roles include serving as a Judge for the Edison Awards, where he was assigned to evaluate approximately 90 nominations in technical categories, and confirmed participation as a Keynote Speaker at various international conferences. His status is also affirmed by serving as a Judge for the BIG Awards for Business.
• Featured Presentations and Authorship: His expertise has led to completed presentations at major scientific forums, including presenting his research at ICECST 2025 (IEEE-associated) and ICDPN 2025 (Springer). These events confirm his expertise in highly technical domains and his status as a published researcher in association with major publishers.
• Distinguished Fellowships: His status as an IFGICT Fellow is secured, and he has received an official invitation to join the SCRS Distinguished Fellow community, which signals recognition for his substantial contributions in Data Science, AI, and Healthcare Informatics.
• IEEE Peer Review: His expertise is also utilized in the rigorous academic evaluation process as a peer reviewer for IEEE publications and conferences, where his ability to critically assess complex research methodology is necessary for maintaining scholarly standards.

By relentlessly focusing on the Verifiable AI Ecosystem, Tirumalasetty is building the framework of trust that ensures the technologies designed to save lives are inherently safe, fully auditable, and reliable.