For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

LoginCreate account
Invite to Job
Piyumi Prasadini

Piyumi Prasadini

AI/Machine Learning Engineer

SRI_LANKA flag
N/A, Sri Lanka
$10.00/hrIntermediateOtherLabel Studio

Key Skills

Software

Other
Label StudioLabel Studio

Top Subject Matter

Regulatory Decision-Making
ML
AI/NLP

Top Data Types

TextText
ImageImage
AudioAudio

Top Task Types

Emotion Recognition

Freelancer Overview

Machine Learning Engineer—LLM Data Collection and Fine-tuning. Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal, Proprietary Tooling, and Other. Education includes Bachelor of Science in Engineering with Honours, University of Ruhuna. AI-training focus includes data types such as Text, Image, and Audio and labeling workflows including Fine-tuning, Evaluation, and Rating.

IntermediateEnglish

Labeling Experience

Developer—Audio Labeling for Voice Emotion Analyzer

AudioEmotion Recognition
Developed an audio-based emotion detection system using speech data and AI models. Labeled and annotated speech audio clips with emotion categories for model training and validation. Used Librosa and PyTorch for preprocessing, feature extraction, and annotation. • Created labeled datasets mapping audio clips to emotional states • Worked with pre-trained Wav2Vec 2.0 models on labeled data • Performed manual validation and adjustment of emotion labels • Applied audio analysis and filtering for high-quality annotation

Developed an audio-based emotion detection system using speech data and AI models. Labeled and annotated speech audio clips with emotion categories for model training and validation. Used Librosa and PyTorch for preprocessing, feature extraction, and annotation. • Created labeled datasets mapping audio clips to emotional states • Worked with pre-trained Wav2Vec 2.0 models on labeled data • Performed manual validation and adjustment of emotion labels • Applied audio analysis and filtering for high-quality annotation

2025 - 2025

Developer—Labeling and Classification for Sketch-to-Color Generator

ImageClassification
Designed and trained a deep learning model to colorize hand-drawn sketches from grayscale to full color. Implemented and managed data preprocessing and image classification pipeline during project. Used TensorFlow and OpenCV tools for managing and labeling training images. • Labeled input images with ground-truth color outputs for model learning • Performed quality control for the labeled datasets before training • Automated preprocessing label assignment via image transformation scripts • Oversaw iterative model refining based on labeled validation batches

Designed and trained a deep learning model to colorize hand-drawn sketches from grayscale to full color. Implemented and managed data preprocessing and image classification pipeline during project. Used TensorFlow and OpenCV tools for managing and labeling training images. • Labeled input images with ground-truth color outputs for model learning • Performed quality control for the labeled datasets before training • Automated preprocessing label assignment via image transformation scripts • Oversaw iterative model refining based on labeled validation batches

2025 - 2025

AI Developer—Evaluation and Labeling for AI-based Depression Detection

OtherText
Developed and evaluated a multi-agent AI system for depression detection and therapy using large language models and agentic AI approaches. Conducted real-world testing to assess model classification accuracy and therapy recommendation effectiveness. Used BERTScore-based comparisons to select optimal AI model workflows. • Annotated and labeled classification outcomes during system evaluation • Focused on improving depression-level classification accuracy via label validation • Implemented RAG and prompt-based evaluation loops • Leveraged LangChain and Python in labeling and evaluation stages

Developed and evaluated a multi-agent AI system for depression detection and therapy using large language models and agentic AI approaches. Conducted real-world testing to assess model classification accuracy and therapy recommendation effectiveness. Used BERTScore-based comparisons to select optimal AI model workflows. • Annotated and labeled classification outcomes during system evaluation • Focused on improving depression-level classification accuracy via label validation • Implemented RAG and prompt-based evaluation loops • Leveraged LangChain and Python in labeling and evaluation stages

2025 - 2025

Machine Learning Engineer—LLM Data Collection and Fine-tuning

TextFine Tuning
Led the data collection team for a RAG-based LLM project, overseeing acquisition and curation of textual datasets for AI model improvement. Contributed to the fine-tuning and development of language models to enhance regulatory decision-making. Performed iterative training and evaluation of LLM outputs with zero-shot prompting techniques. • Responsible for ensuring high-quality, diverse data samples for LLM fine-tuning • Applied advanced prompt engineering strategies for tailored regulatory scenarios • Utilized LangChain, PyTorch, and proprietary tools in the workflow • Coordinated team efforts to maintain annotation consistency and data validation

Led the data collection team for a RAG-based LLM project, overseeing acquisition and curation of textual datasets for AI model improvement. Contributed to the fine-tuning and development of language models to enhance regulatory decision-making. Performed iterative training and evaluation of LLM outputs with zero-shot prompting techniques. • Responsible for ensuring high-quality, diverse data samples for LLM fine-tuning • Applied advanced prompt engineering strategies for tailored regulatory scenarios • Utilized LangChain, PyTorch, and proprietary tools in the workflow • Coordinated team efforts to maintain annotation consistency and data validation

2024 - 2024

Education

U

University of Ruhuna

Bachelor of Science in Engineering with Honours, Computer Engineering

Bachelor of Science in Engineering with Honours
2021

Work History

V

VSIS

Data Analyst & Visualization Engineer Intern

N/A
2024 - Present
O

Omdena

Machine Learning Engineer

N/A
2024 - 2024