Ashutosh Adhikari | ML Engineer

Professional Profile

I specialize in building machine learning solutions with expertise in NLP, computer vision, and data engineering. My work focuses on creating scalable ML pipelines and implementing cutting-edge AI research in production environments.

Core Expertise:

AWS SageMaker for model training and inference
ML pipeline development and optimization
Natural Language Processing (NLP)
Data visualization with React and Sigma.js
Explainable AI (XAI) techniques

Work Experience

TAI Inc. Jan 2022 - Present

ML Engineer

Developed ML-based search application for Nomura Research Institute (Japan's top 10 fintech company)
Collaborated with AMBL (AI Research Company in Japan) to deliver production-ready solutions
Built common representation models for large Japanese datasets
Implemented AWS SageMaker workflows for scalable model training and inference
Created data visualization tools using React and Sigma.js for graph databases
Researched and implemented XAI techniques for NLP model interpretability
Mentored and guided interns in professional development

Python AWS SageMaker PyTorch React Neo4j

TAI Inc. Sep 2021 - Dec 2021

ML Intern

Researched ML applications in music generation and synthesis
Implemented state-of-the-art AI techniques for music projects

Python Magenta.js TensorFlow

CloudFactory Jul 2020 - Aug 2021

Annotation QA

Ensured quality of annotated image data for ML datasets
Developed validation processes for labeling accuracy

Data Annotation Quality Assurance

Technical Skills

Programming Languages

Python (Advanced)
JavaScript (Intermediate)
React (Intermediate)

Databases

MySQL
Neo4j
Elasticsearch

ML Frameworks

PyTorch
TensorFlow
Scikit-learn
SageMaker SDK
Langchain

Tools & Technologies

AWS SageMaker
Docker
Jenkins
GitLab CI/CD
Pandas
Jupyter
Sigma.js

Featured Projects

Aslead Search

Python PyTorch NLP

Code classification tool for technical documents achieving 95% accuracy with NLP models.

Implemented Hugging Face transformer models for document classification
Utilized XAI techniques for BERT model visualization
Built training/inference pipeline with SageMaker SDK

Relative Analyzer

React Sigma.js Neo4j

Web application for visualizing data stored in Neo4j graph databases.

Optimized Neo4j queries (50x performance improvement)
Integrated ElasticSearch for enhanced search capabilities
Developed Docker-based deployment pipeline

Melody Synthesis

Python Magenta.js TensorFlow

ML-powered system for generating new musical melodies.

Implemented state-of-the-art music generation models
Developed user-friendly interface for music synthesis
Integrated ML models with web application

Education

The British College

Bachelor of Science in Information Technology (BSc IT)

2017 - 2021

Relevant Courses:

Data Structures
Discrete Systems
Cognitive Science
Artificial Intelligence
Database Administration

Get In Touch

Contact Information

[email protected]

+977 98XXXXXXX

Kathmandu, Nepal