Avyesh B.

Key Skills

Software

Other

Top Subject Matter

Medical AI

Clinical Diagnostics

Facial Affect Recognition

Top Data Types

Audio

Image

Video

Top Task Types

Transcription

Emotion Recognition

Classification

Diagnosis

Prompt + Response Writing (SFT)

Evaluation/Rating

Question Answering

Freelancer Overview

I am a 4th-year MD student at the Medical University of Warsaw and a B.S. graduate in Electrical & Computer Engineering from the University of Arizona USA. A combination that very few applicants bring to this field. While most candidates come from either medicine or engineering, I work fluently across both: I have built a fully functional multimodal medical AI system called DR.AI from scratch (integrating audio transcription, facial affect analysis, and gait recognition for clinical assessment), while simultaneously conducting grant-funded biomedical research in biomarker validation using ELISA, FibroScan, and Python-based ML pipelines. I have 4 peer-reviewed publications, founded WUM's plastic surgery student club, and have hands-on surgical experience across plastic & reconstructive surgery, orthopedics, transplant surgery, and general surgery. I am both a medical student who is curious about technology, and an engineer who knows his Medicine, rare intersection creates real value.

Labeling Experience

ML Pipeline Contributor – Medical Data Labeling for Biomarker Validation

OtherDiagnosis

Curated, preprocessed, and labeled clinical data—including ELISA, FibroScan, and histology images—for ML-based biomarker validation pipelines. Applied standardized diagnostic and classification labels to medical datasets to facilitate supervised machine learning. Used Python tools to extract features, annotate datasets, and manage large multi-modal medical data. • Labeled ELISA and FibroScan results for disease classification • Managed dataset integrity and structured clinical metadata annotation • Assisted in validation of diagnostic ML models via labeled data • Developed scripts for automated preprocessing and labeling tasks

2025 - Present

AI Platform Developer – Gait Video Annotation and Classification

OtherVideoClassification

Developed and annotated gait analysis datasets for use in clinical AI assessment of patient mobility. Classified video segments according to standardized gait evaluation criteria and labeled key events and patterns. Managed dataset organization and quality assurance throughout the labeling process. • Employed Python and custom scripts for video segmentation and labeling • Set up annotation protocols for gait classification • Ensured rigorous quality standards for training data • Facilitated integration of labeled datasets into AI clinical toolchain

2025 - 2026

AI Platform Developer – Facial Emotion Annotation

OtherImageEmotion Recognition

Built facial affect recognition modules using MediaPipe and OpenCV as part of a unified medical AI platform. Labeled and annotated clinical facial images to capture emotion and affective state for AI model training. Designed systematic annotation procedures to ensure high inter-annotator reliability and dataset consistency. • Utilized MediaPipe and OpenCV for landmark detection and labeling • Created protocols for facial affect annotation • Structured dataset documentation and metadata curation • Performed quality control on annotated facial image data

2025 - 2026

AI Platform Developer – Clinical Audio Transcription & Annotation

OtherAudioTranscription

Led development of a multimodal clinical AI platform integrating audio, facial, and gait data for automated exam room documentation. Conducted extensive audio transcription using Whisper to create labeled datasets for training clinical speech recognition models. Implemented preprocessing and structured annotation of audio segments reflecting clinical conversations. • Developed workflow for audio data ingestion and labeling • Ensured data quality through manual verification and cleaning • Utilized Whisper and Python scripting for annotation • Evaluated model performance using labeled audio datasets

2025 - 2026

Education

Work History