Founder's Insight · July 2024 · 5 min read
The Future of AI in Healthcare: Key Findings From My Journey as a Technologist & Founder
AI is no longer a distant promise for healthcare — it is becoming the backbone of a new global health ecosystem. Drawing on hands-on work across medical imaging, assistive robotics, and pharma compliance, this piece covers the biggest insights shaping the field: from why edge AI matters more than cloud AI in clinical settings, to why human oversight isn't a constraint on AI ambition but the foundation of it. Written by our founder after years of building in this space.
July 29, 2024 · Founder's Perspective
Case Study · Edge AI
Netra Sakhi: building an eye-care assistant for the edge
Netra Sakhi is a virtual eye-care assistant we designed to run as a real-time edge AI system — bringing vision-screening support to places where specialist access is limited. Published as a Scopus-indexed system in 2023, it became the first deployed proof that Praverse's healthcare AI could work outside the lab: low-latency inference on constrained hardware, a conversational front-end, and a screening pipeline tuned for accessibility rather than a research benchmark. The lessons from Netra Sakhi — privacy-first design, edge inference, and a human-in-the-loop screening flow — now run through everything HealthMate does.
2023 · Edge AI / Healthcare
Vision · HealthMate
Why HealthMate listens before it answers
HealthMate started from a simple observation: most patients don't need a diagnosis at the front door — they need someone to listen, route them to the right place, and hand clinicians a clear picture of what's going on. HealthMate is built as a voice-first companion for intake, hospital navigation, appointment assistance, and clinician-ready summaries, with humanoid and kiosk deployment explored through digital-twin simulation before anything touches the physical world. It is patent-pending, and designed to support — never replace — medical professionals.
2025 · Patent-Pending · Coming Soon
Explainer · Pharma AI
What "AI-assisted" actually means in a GMP environment
In regulated industries, an AI system that hallucinates a compliance answer is worse than no AI at all. That's why Praverse's pharma and regulatory intelligence systems are built around evidence: every FDA-483 observation, CAPA suggestion, or data-integrity flag links back to the source document it came from, with a full audit trail of what the system surfaced and what the expert decided. The AI drafts, analyzes, and flags — quality professionals decide. That hierarchy isn't a policy on top of the system; it's built into the architecture itself.
Ongoing · Regulatory Intelligence
Research Note · Imaging
From DME detection to PCOS screening: one imaging pipeline, two problems
Our published work on Diabetic Macular Edema (DME) detection (IEEE Transactions on Medical Imaging, 2023) and the PCOS detection pipeline (2024) share a common foundation — deep learning models trained for diagnostic-support, not diagnosis. Both pipelines are designed to flag patterns for expert review rather than issue verdicts, and both feed directly into the Medical Imaging Lab's ongoing work on fundus and ultrasound screening tools that can run at the edge.
2023–2024 · Medical Imaging Lab
