Tirop Meshack

mtirop345@gmail.com +82 109 672 2363 Goseong, Gangwon-do, South Korea github.com/meshtirop1 linkedin.com/in/meshack-tirop

Career objective

To build a career at the intersection of machine learning, bioinformatics, and full-stack software development. I aim to apply deep learning and data-driven techniques to real-world healthcare challenges, particularly cardiovascular and neurocritical care, while continuing to grow as a developer across web and mobile platforms. In the long term, I want to contribute to open-source research tooling and build applications that improve lives through intelligent systems.

Education

BSc, Computer Engineering (Smart Computing) 2022 – 2026 (expected)

Kyungdong University, Goseong, Gangwon-do, South Korea

Focus: web development, bioinformatics, machine learning, deep learning.

Senior thesis: Temporal prediction of heart defects (CheXchoNet, CNN, AUC 0.79–0.84). Supervisor: Prof. Zubaer Ibna Mannan.

Relevant coursework: Machine Learning · Deep Learning · Bioinformatics · Data Structures · Web Development.

Kenya Certificate of Secondary Education 2017 – 2021

Kaplelach High School, Kenya

Technical skills

Languages: Python · JavaScript · HTML · CSS · SQL · Solidity (learning)

Web & backend: Django · Django REST Framework · PostgreSQL · Git · Jupyter

Mobile: React Native · Expo · OpenGraphs

Ops & observability: AWS · Sentry · Grafana

ML & AI: PyTorch · TensorFlow · Scikit-learn · Pandas · NumPy · PyCharm

ML techniques: Supervised / unsupervised learning · CNNs · LSTMs · TCNs · Transformers · temporal & waveform prediction · feature engineering

Databases: PostgreSQL · SQL

Other: OOP · data structures · technical writing · presentations

Research

Senior Thesis: Temporal Prediction of Heart Defects (CheXchoNet) 2023 – 2024

Senior thesis · Kyungdong University · Supervisor: Prof. Z. I. Mannan

Deep-learning model for temporal prediction of heart defects on the CheXchoNet dataset.
Used CNNs to analyse chest X-ray data alongside temporal metadata, rather than treating each frame in isolation.
Achieved AUC 0.79 to 0.84, demonstrating robust performance for clinical-imaging applications.
Pointed me toward waveform-level clinical signals, which the current ICP research builds on.

Current Research: Non-invasive ICP Estimation 2025 – Present

Ongoing research · Kyungdong University · Supervisor: Prof. Z. I. Mannan

Training LSTM (bidirectional, attention), TCN, Transformer, and Hybrid (CNN+Transformer+autoreg) models on ABP and CBFV waveforms (10 s windows, 125 Hz).
Trained on PhysioNet's pediatric neurocritical-care waveform release; cross-dataset validation against PhysioNet CHARIS (adult ICP).
Leave-one-patient-out (LOPO) cross-validation across 12 patients, ~18,500 windows.
Best LSTM result: MAE 2.92 mmHg, median AE 1.97 mmHg, RMSE 4.03 mmHg.
Stack: PyTorch, NumPy, Pandas, SciPy, custom WFDB / signal-processing pipeline.

Cardiovascular Disease Prediction 2023

Kyungdong University · Cleveland Clinic dataset

Trained a machine-learning model to predict cardiovascular disease.
Implemented feature engineering and model-optimisation techniques to improve prediction performance.
Achieved 96% test accuracy using Scikit-learn and PyTorch.

Professional experience

Web Developer, Sunrise Drilling Limited Dec 2024 – Present

Freelance · Kenya · Remote

Maintain and extend the corporate site on Django and PostgreSQL with a secure admin backend.
Set up Grafana dashboards for traffic and uptime; tightened deploy and on-content-update workflows.
Working stack: SQL, JavaScript, Django, PostgreSQL, Grafana.

Backend Developer, Multition Education Jun 2024 – Present

Freelance · Kenya · Remote

Backend developer on multition.com, an AI-focused education platform with mentor-guided courses across ML, DL, NLP, computer vision, and generative AI. 500+ students, 11+ courses, 50+ certificates issued.
Built and maintained Django services for cohort enrolment, course content, and certificate generation.
Instrumented the backend with Sentry for error monitoring and alerting.
Working stack: Python, Django, PostgreSQL, Sentry, AWS.

Independent Freelance, Full-Stack Web & Mobile 2022 – Present

Independent · clients in Kenya & South Korea

Designed and built responsive websites and mobile apps using Django, React Native, HTML, CSS, and JavaScript.
Shipped jiranisoko.co.ke, a Kenyan neighbourhood marketplace with listings, jobs, motors, property, and M-Pesa integration.
Shipped teddyvisuals.com, a creative portfolio and media site.
Built a service-business site with custom request flow for Mamafua Agency.
Implemented PostgreSQL databases for data management and user authentication.
Developed cross-platform mobile apps using React Native, Expo, and OpenGraphs.

Certifications & training

Smart Contracts, Cyfrin Updraft (Nov 2024 – Mar 2025).
Data or Specimens Only Research, CITI Program (Jan 2025; ID 67095047; expires Jan 2028).
Conflicts of Interest, CITI Program (Jan 2025; ID 67074050; expires Jan 2029).
Python for Beginners, Simplilearn (Nov 2022).

Research access

Granted credentialed access to PhysioNet's sensitive datasets, including Neurocritical Care Waveform Recordings in Pediatric Patients (1.0.0) and CHARIS, both used in the current ICP work.

Memberships

IEEE (Institute of Electrical and Electronics Engineers), active member.

Languages

English: Fluent

Swahili: Native

Korean: Beginner

Professional interests

Bioinformatics research, especially cardiovascular and neurocritical signals.
Mobile and web application development.
Open-source contributions.
Machine learning in healthcare.

References

Prof. Zubaer Ibna Mannan zimannan@kduniv.ac.kr

Assistant Professor, Smart Computing, Kyungdong University · Bioinformatics & thesis supervisor

Dr. Baseem Al-Athwari bassem_cs@kduniv.ac.kr

Associate Professor & Head, Smart Computing, Kyungdong University

Curriculum Vitae.