About me

I am Sonika Acharya, a Machine Learning Engineer passionate about using AI to solve real-world challenges and create innovative solutions. My journey in tech has been driven by hands-on experience with advanced machine learning models and an eagerness to push the boundaries of what's possible. I love exploring new technologies and staying engaged with the latest trends in AI. I am always eager to learn about new developments and discoveries in the field, whether through industry events, online courses, or reading recent research papers. Staying engaged with the evolving landscape of AI helps me stay inspired and brings fresh ideas to my work.

In my free time, I enjoy diving into Nepali literature, which keeps me connected to my cultural roots and sparks my creativity. I love traveling and exploring new places, which offers me a refreshing escape from the routine and brings new experiences and perspectives that inspire me.

Projects

Image Caption Generator

Chatbot Application

Eye Gaze Detection

Speech Emotions Recognition

Image Cartoonization

Resume Parser

Work

Aug, 2023 - Present

Machine Learning Engineer - Fusemachines Inc.

Preprocessed and formatted data to optimize input and fine-tuned LayoutLM model to extract information from legal documents.

Worked on an end-to-end pipeline for financial document information extraction, integrating DETR and DiT models. Post-processing included regex for pattern extraction and LLMs for context-aware extraction.

Researched various papers on arbitrary shaped text detection and evaluated findings based on accuracy and limitations to determine the best architecture for implementation.

Engaged in knowledge-sharing sessions, conducting and reviewing papers on topics such as Explainable AI and Model Quantization.

Dec, 2022 - Mar, 2023

Machine Learning Engineer (Intern) - TAI Inc.

Developed the TAIChat API endpoint, seamlessly integrated into the web project, facilitating general conversation and providing comprehensive information about TAI.

Developed an application for eye gaze detection as a part of Computer Vision project.