Real projects, delivered inside a working hospital
From AI models running in critical care to multi-crore procurement decisions and accreditation systems — here's a selection of what I've actually built and delivered.
A closer look at three signature projects
The challenge, the approach, and the result — the work behind the headlines.
Ventilator Failure Prediction
Critical-care ventilators were maintained reactively — problems surfaced only when a machine failed, putting patients at risk and keeping the biomedical team in constant firefighting mode.
I built a Random Forest machine-learning model monitoring 17 parameters across 22 ventilators in ICU, ICCU, ER and Paediatrics, generating an ongoing failure-risk score for each unit.
Critical-care maintenance shifted from reactive to predictive. The project was presented to and endorsed by hospital leadership, with a roadmap to extend it to anaesthesia and dialysis machines.
ER Triage Chatbot
Emergency-room staff needed fast, accurate answers to SOP and triage questions, and the department needed a stronger case to justify new equipment investment.
I built a Chatbase-powered chatbot trained on the ER's live SOPs, giving staff round-the-clock instant guidance — and used its real-world impact to build a data-backed business case.
Faster SOP access for ER staff, and a business case strong enough to secure management approval for a multi-lakh equipment purchase.
AI-Assisted CAPEX Prioritisation
Leadership faced a multi-crore capital proposal with no objective way to decide what to fund now versus what could safely wait.
I ran an AI-assisted review, triaging every line item into approve-now, conditional and defer tiers — each one mapped to JCI Facility Management & Safety standards.
A defensible, evidence-based investment plan that leadership could sign off with confidence and defend to the board.
Across AI, procurement, compliance, apps and training
Patient-Inflow Prediction Model
A forecasting model to anticipate patient volumes and help plan staffing and resources ahead of demand.
12-Year Cath Lab Data Analysis
A longitudinal procedure-data study that pinpointed the true driver behind a utilisation gap, grounding a major decision in evidence.
Multilingual Triage Support
AI-assisted, multilingual guidance to make emergency information more accessible to diverse patients.
MRI & LINAC Vendor Comparison
Weighted, vendor-neutral evaluation tools for high-value imaging and radiotherapy purchases, with India-specific regulatory checks.
CT Scanner Evaluation
Independent scoring of competing OEMs across clinical, technical and service criteria to support an objective decision.
Equipment Lifecycle Planning
Replacement and condemnation frameworks built on data rather than guesswork or vendor pressure.
AHU Failure RCA & CAPA
A root-cause analysis and corrective-action package for an air-handling unit failure, aligned to NABH FMS requirements.
FMEA & Risk Register Framework
Reusable risk-assessment templates — FMEA, risk register, action tracker and SLA formats — built for NABH/JCI.
MvPI Training & Reporting System
A training programme and reporting workbook for monthly materiovigilance submissions.
Equipment Utilisation Report App
A React app with role-based access and audit trail to track and report medical-equipment utilisation for NABH/JCI compliance.
PPM Management Prototype
A preventive-maintenance scheduling prototype to streamline and document equipment upkeep.
Auto-Reminder Mail System
An automated reminder system built independently with Google Sheets and Apps Script — no external tools required.
AI Workshops
Hands-on AI and ChatGPT workshops delivered to healthcare and corporate teams, with audience-specific use cases.
AI Tools for Healthcare Professionals
My first book — a practical, India-focused guide to AI in clinical and hospital work — now live on Amazon Kindle, with the Biomedical Intelligence Series following.
Healthcare Blog & Education
Patient-facing and technical content that builds trust and explains complex services simply.
Want results like these for your hospital?
Let's talk about what's possible — whether it's AI, accreditation, procurement or a custom tool.
Start a Conversation