Jinwoo Seol
AI Software Engineer
Seoul
Summary
“Applied AI Engineer building production-ready systems”
Product-oriented AI Project Leader & Software Engineer with over 3 years of professional experience building production-ready Bio-IT systems. Currently spearheading the 'Canopus SW' (MSI Diagnostic) project, I bridge the gap between clinical stakeholders and scalable AI architectures by translating complex medical requirements into high-performance software. I specialize in developing end-to-end diagnostic pipelines spanning signal processing, feature extraction, and AI inference to transform high-volume biological data into actionable medical insights.
My work delivers measurable impact, including >0.99 AUC for MSI prediction models and >99.5% accuracy for CGT cell recognition. I have pioneered patented algorithms that improve signal resolution by 2-10 times compared to existing methods. Beyond algorithm development, I ensure seamless production deployment through robust C#–Python integration layers and strict adherence to global regulatory standards such as CE-IVDR and MFDS.
I approach AI development as a product-driven discipline, collaborating directly with pathologists and medical professionals to define MVP requirements and design intuitive UI/UX scenarios. My core mission is to deliver cohesive, secure, and reproducible software systems where advanced algorithms and stable data pipelines work in harmony to solve real-world clinical and business challenges.
Core Competencies
- High-Precision AI Modeling – Experience developing MSI prediction AI with over 95% accuracy on clinical data and deep-learning–based cell recognition models with over 99.5% accuracy in the CGT field
- Digital Analysis Performance Innovation – Development of Digital Melting signal correction and ultra-precise algorithms improving resolution by 2–5 times, with related patents
- End-to-End Development – End-to-end development of diagnostic software from planning to deployment using C/C++, Python, Java, and Qt
- Regulatory & Security Compliance – Experience responding to CE-IVDR/MFDS regulations and implementing cybersecurity functions such as communication encryption, database access control, and data encryption
Overview
3
3
years of professional experience
Timeline
AI Engineer – Digital PCR & Melting Analysis
OPTOLANE
10.2024 - Current
AI Cell Recognition System Development
OPTOLANE
06.2024 - Current
Automated DNA Extraction Device Development
OPTOLANE
09.2023 - Current
Diagnostic PCR Instrument Software Development
OPTOLANE
10.2022 - Current
Deep Learning-Based Breast Cancer Targeting Model
Personal Project
03.2022 - 06.2022
Bachelor of Science - Double Major: Life Science & Computer Engineering
POSTECH (Pohang University of Science And Technology)
Work History
AI Engineer – Digital PCR & Melting Analysis
OPTOLANE
10.2024 - Current
- End-to-End Product Leadership: Orchestrated the full development lifecycle of 'Canopus SW', leading project planning, MVP definition, and system architecture based on clinical requirements gathered from pathologists.
- High-Impact AI Engineering: Constructed an end-to-end analysis pipeline achieving >0.99 AUC on clinical MSI data, integrating ML-based QC modules for melting curve separation and noise removal.
- Patented Technical Innovation: Designed and patented core ultra-precise algorithms that improved data resolution by 2–10 times, significantly outperforming existing diagnostic methods.
- Scalable System Integration: Engineered a robust C#–Python integration layer to link Python-based AI algorithms with a production-ready application environment, ensuring high-performance execution.
Tech: Python, C#, Scikit-learn, Cursor AI
AI Cell Recognition System Development
OPTOLANE
06.2024 - Current
- Design of a CGT analysis system combining AI-based cell recognition with proprietary PCR assay data
- Development of a deep-learning–based cell recognition model with over 99.5% accuracy for automated microscopy image analysis
- Construction of the full pipeline for CGT analysis including data preprocessing, AI inference, and result visualization
- Independent development of a prototype including UI and analysis functions within two months using Cursor AI
- Tech: PyTorch, TensorFlow, PySide6, PyQt, Cursor AI
Automated DNA Extraction Device Development
OPTOLANE
09.2023 - Current
- Completion of a prototype through establishment of core functions and architecture of a nucleic acid extraction system
- Design of UI scenarios and system optimization aimed at usability and security based on analysis of medical professionals’ requirements
- Development of automated consumable inspection functions and optimal path calculation systems capable of flexibly responding to various situations
- Reduction of the device control interface development period from four months to two months through outsourcing vendor management, shortening the overall project schedule
- Preparation of user manuals and technical documents for CE and MFDS certification
- Tech: StarUML, Draw.io, Android, Java
Diagnostic PCR Instrument Software Development
OPTOLANE
10.2022 - Current
- Development of Qt-based UI and analysis algorithms on Linux and Windows platforms
- Development of diagnostic data integration functions with AWS servers using Java on Windows
- Diagnostic data management and performance optimization using MariaDB
- Design and implementation of cybersecurity functions required for medical device software, including equipment and network communication encryption, database access control, and stored data encryption
- Compliance with regulation-based development processes through preparation of user manuals and technical documentation for CE-IVDR and MFDS certification
- Tech: C++, Qt, Java, Linux, MariaDB, SQL, Eclipse, SVN
Deep Learning-Based Breast Cancer Targeting Model
Personal Project
03.2022 - 06.2022
- Design of a breast cancer prediction model based on EfficientNet-B2, achieving an initial accuracy of 86%
- Improvement of model accuracy up to 93% by applying probability-density–based targeting enhancement techniques
- Tech : NumPy, Pandas, Matplotlib, Seaborn, TensorFlow, Keras, Scikit-learn, Jupyter Notebook
Education
Bachelor of Science - Double Major: Life Science & Computer Engineering
POSTECH (Pohang University of Science And Technology)
South Korea02.2022
Skills
Deep Learning, Computer Vision
PyTorch, TensorFlow, Sklearn
Python, C/C, Java
Qt / PySide6 UI Development
MariaDB & Secure DB Control
Secure Communication Protocols
Linux / Windows
Git / SVN
Medical Development Experience
Neurogenetics Laboratory, POSTECH
- Modeling of neurodevelopmental disorders associated with congenital brain malformations.
Neuro-Epigenetics Laboratory, POSTECH
- Characterization of activity-dependent transcriptional regulation of the Arc gene in neuronal cells.
Molecular Neuroscience Laboratory, POSTECH
- Viral-mediated neural pathway mapping for anatomical and functional connectivity confirmation.
Analysis Team, Genexine, Inc.
- Development and validation of DNA-based therapeutic vaccine constructs targeting lung cancer.
C-BIOMEX
- Development of peptide-based diagnostic biomarkers for breast cancer and assessment of peptide–protein binding affinity.
Korea Brain Research Institute (KBRI)
- Analysis of behavioral phenotypes and gene-expression alterations in murine depression models under neural stimulation.
Medical Development Experience
Neurogenetics Laboratory, POSTECH
- Modeling of neurodevelopmental disorders associated with congenital brain malformations.
Neuro-Epigenetics Laboratory, POSTECH
- Characterization of activity-dependent transcriptional regulation of the Arc gene in neuronal cells.
Molecular Neuroscience Laboratory, POSTECH
- Viral-mediated neural pathway mapping for anatomical and functional connectivity confirmation.
Analysis Team, Genexine, Inc.
- Development and validation of DNA-based therapeutic vaccine constructs targeting lung cancer.
C-BIOMEX
- Development of peptide-based diagnostic biomarkers for breast cancer and assessment of peptide–protein binding affinity.
Korea Brain Research Institute (KBRI)
- Analysis of behavioral phenotypes and gene-expression alterations in murine depression models under neural stimulation.