- AI/NLP Engineer specializing in LLM fine-tuning, transformer-based NLP, and RAG-optimized knowledge pipelines, with hands-on experience building agentic LLM systems and domain-focused Custom GPT workflows. Builds and automates intelligent, end-to-end workflows using multi-agent architectures with n8n, LangChain, and Model Context Protocol (MCP), leveraging LangGraph for stateful execution and LangSmith for tracing, evaluation, and observability. Experienced in deploying production-grade ML and LLM workflows using GitHub Actions, Docker, Kubernetes, MLflow, DVC, and AWS, delivering reliable, automated, and observable enterprise AI solutions.
Pinned Loading
-
Email_Multiclass_classification_LSTM
Email_Multiclass_classification_LSTM PublicThe built model(LSTM) is able to Categorize emails into Primary, Spam, Promotions and Social. Because of dataset taken from individual, it's not uploaded for privacy concern.
Jupyter Notebook
-
MLOps_text_classification_project
MLOps_text_classification_project PublicFully Deployed Text classification based app on AWS EKS with MLflow (DAGSHub), DVC + S3, Docker, GitHub Actions CI/CD, Prometheus & Grafana
Python
-
MLOps_vehicle_insurance_project
MLOps_vehicle_insurance_project PublicA fully-automated, end-to-end MLOps project that includes data ingestion from MongoDB, data validation, data transformation, model training, model evaluation, CI/CD pipelines, Dockerization, AWS de…
Python
-
romanized-nepali-text-sentiment-analysis
romanized-nepali-text-sentiment-analysis PublicThe built model (Naive Bayes: Multinomial, Gaussian and Bernoulli combined voted) is Sentiment Analyzer for Romanized Nepali Text (Nepali words spoken in English Letters).The model predict either p…
Jupyter Notebook
If the problem persists, check the GitHub status page or contact support.
