Our story

Born in a research lab.
Hardened in industry.
Built for everyone in between.

Logic Byte exists because the gap between powerful imaging instruments and the insights they contain is wider than it should be — and bridging that gap shouldn't require a PhD in image processing.

2019

The problem we saw every day

The idea for Logic Byte was born in a microscopy lab. During years of research at the National University of Singapore, one pattern kept repeating: the scientists gathering and using imaging data were world-class in their fields — cell biology, pathology, materials science, semiconductor fabrication — but they were not experts in image processing, digital signal processing, or programming.

They were spending hours on manual analysis, relying on brittle scripts, or accepting lower-quality results simply because better tools were out of reach. Powerful algorithms existed in the research literature, but the gap between a published method and a usable, reproducible workflow was enormous.

"The bottleneck wasn't the instrument. It wasn't even the data. It was the processing pipeline between raw image and real insight."

The value of an imaging experiment — and the decisions it enables — is determined almost entirely by how well the images are processed and analysed. That felt like the right problem to solve.

2021

Logic Byte is founded

Logic Byte was founded in 2021 and incorporated in Singapore in 2022, drawing on extensive experience in autonomous microscopy, deep learning, and computer vision built during a PhD and postdoctoral research career.

The core idea was straightforward: build a customisable, visual workflow builder that puts sophisticated, domain-aware image analytics in the hands of the domain expert — without requiring them to write a single line of code. The platform would handle the domain-aware pipeline; the expert would focus on the science.

We started with the fundamentals and added capabilities methodically, each one validated against real imaging problems:

Image enhancement & correction Dynamic range adjustment, illumination correction, and denoising — the essential first step before any downstream analysis.
Object detection & segmentation Automated detection and precise delineation of cells, structures, and defects across diverse imaging modalities.
Morphological analysis & data extraction Per-object measurements, population statistics, and structured data export — turning pixels into interpretable numbers.
Single-cell analysis High-content analysis of individual cells: classification, phenotypic scoring, spatial mapping, and clustering.
Defect detection Supervised and unsupervised inspection modules for semiconductor wafers, PCBs, and industrial surfaces — same platform, different domain.
Label-free in-silico staining (CytoByte) Our most recent breakthrough: rendering pseudo-DAPI nuclear staining from standard DIC microscopy images using deep neural networks — no chemicals, no preparation, no additional equipment.
Now

A platform that grows with you

Today, OptiFlow is a mature, modular image analytics platform that spans two primary verticals — Biomedical & Life Sciences and Semiconductor & Industrial Inspection — with a custom solutions track for organisations with unique imaging challenges.

The same workflow builder, the same scalable deployment architecture, the same commitment to reproducibility — configured for your domain. A BioPharma team running drug screening assays and a semiconductor fab running wafer inspection are solving very different problems, but they share the same need: reliable, auditable, scalable image analysis that does not require a dedicated programming team.

OptiFlow workflows run on a single research workstation or deploy at scale across an entire production line. Results are fully reproducible — the same pipeline, the same parameters, the same outputs, every time.

What drives us

Every field that relies on imaging — pathology, drug discovery, materials science, semiconductor manufacturing, cell biology — is sitting on a reservoir of untapped insight. The data is there. The instruments are there. What's missing is the analytical layer that makes it accessible.

"From the lab bench to the production floor — analytics built by those who've worked on both."

Logic Byte's mission is to close that gap: to make production-grade, research-validated image analytics available to any team with images to analyse, regardless of their programming background. We believe the next wave of discoveries — in medicine, in manufacturing, in science — will come from people who can finally see what their images are telling them.

The founder
Doorgesh Sharma Jokhun, PhD — Founder, Logic Byte
Founder
Doorgesh Sharma Jokhun
PhD · National University of Singapore | AI & Computer Vision

Doorgesh is a computer vision and AI researcher who spent his PhD and postdoctoral years working at the intersection of microscopy, deep learning, and biomedical image analysis. His research at NUS spanned autonomous microscope control, biosensor CNN development, in-silico nuclear staining, and large-scale single-cell analysis — resulting in over 20 peer-reviewed publications in signal processing, deep learning, and biomedical imaging.

Beyond the research lab, Doorgesh has accumulated direct professional engineering experience across two very different, image analytics-intensive domains: complex biomedical image analysis in clinical and research-grade microscopy environments, and high-precision semiconductor image analytics where measurement accuracy and throughput are non-negotiable. This dual-domain background is rare — and it is what gives OptiFlow its unusual breadth without sacrificing domain depth.

The experience of watching domain experts struggle with image processing bottlenecks — in labs and facilities equipped with world-class instruments — was the direct motivation for Logic Byte. He founded the company to translate research-grade imaging AI into tools that working scientists and engineers can actually use.

PhD · NUS 20+ Publications Computer Vision Deep Learning Biomedical Imaging Autonomous Microscopy In-silico Staining Semiconductor Inspection Signal Processing