Raghubansh Gupta

WORK HISTORY

EDUCATION

SKILLS

ABOUT RAGHUBANSH GUPTA

I architect intelligent imaging systems that unify optics, cameras, ISP, microscopy acquisition, color science, computer vision, machine learning, and system software into a coherent, reliable platform. My philosophy is simple: if a system must be trusted, each layer—from photons to pixels to decisions—must be characterized, calibrated, and continuously verified.My work begins at the physical level: characterizing optics, modeling illumination, quantifying sensor behavior, and correcting intrinsic distortions. These components are then calibrated to specification to ensure color accuracy, spatial fidelity, focus stability, and reproducible capture across time and conditions.On this foundation, I build the algorithmic stack across optical correction, reconstruction, enhancement, feature extraction, and CV/ML inference, all designed with fail-safes, confidence checks, fallback paths, and verification routines that prevent silent failure.The system is designed for continuous monitoring. I architect pipelines that track optical drift, sensor and ISP behavior, autofocus stability, stitching accuracy, tissue-coverage consistency, ΔE color drift, and feature-distribution shifts. Each layer includes self-diagnostics, anomaly detection, and closed-loop feedback that triggers recalibration or adapts parameters as conditions evolve.As datasets evolve, slide preparation varies, or operational conditions shift, the system applies triangulation, graph-based consistency checks, and statistical drift analysis to maintain fidelity—operating as an adaptive, self-correcting, closed-loop system rather than a static algorithm.For the past nine years, I’ve applied this architecture to help build and scale the digital-pathology platform that began as Spectral Insights, expanded through nference, and now powers Pramana → Evident Scientific—a global WSI ecosystem that has digitized 25M+ slides with optical rigor, system-software discipline, automation, and operational QA.I am an inventor on multiple US patents in slide acquisition, label-text analysis, color normalization, rescan automation, visualization, and system-level QA. I also help shape Pramana’s patent strategy, mentoring teams in identifying and protecting innovation.Previously at imec, LG Soft India, and Hi-Tech Robotics, I worked on hyperspectral imaging, camera-ISP pipelines, sensor calibration, image-quality optimization, and autonomous driver-monitoring systems—reinforcing my belief that high-performance imaging emerges only when physics, algorithms, and system software evolve together.