Recent
rag_chatbot
RAG chatbot built with FastAPI and LangChain for retrieval-augmented question answering.
Demonstrates practical GenAI system integration using modern LLM tooling and API serving.
Open projectAI Engineer Portfolio
Building reliable LLM and ML systems for real-world use.
I am Neeraj Kumar Singh, an AI engineer with a computer science background from ISI Kolkata. My focus is applied GenAI, model-driven product features, and engineering workflows that move from research prototypes to usable software.
Selected Projects
View all on GitHubRecent
safespeech-indic-hate-mitigation
SafeSpeech pipeline for Indic hate intensity classification, explanation, and mitigation suggestions.
Shows responsible AI workflow design for safety-critical social text moderation tasks.
Open projectRecent
twitter_sentimental_analysis
Sentiment analysis pipeline for topic-based tweet streams using Python and Twitter API data.
Implements full NLP flow from data collection to sentiment classification and reporting.
Open projectRecent
healthcare_fraud_classification
End-to-end healthcare fraud detection pipeline using claims, beneficiary, and provider datasets.
Benchmarks multiple ML models (RF, LightGBM, DT, SVM, ANN) for fraud-risk classification.
Open projectRecent
IC50
Repository for IC50-related experimentation and predictive modeling workflow.
Represents domain-specific ML experimentation and applied model development practices.
Open projectRecent
quiz_app
Quiz application project focused on interactive quiz flows and backend-driven score handling.
Demonstrates full-stack implementation skills for user-oriented assessment products.
Open projectResearch and Technical Focus
LLM application design and prompt-driven product features
Deep learning systems: ANN, CNN, ResNet, and ResNeXt implementation
Backend-heavy ML product delivery with Python and Django
Writing and Talks
A dedicated publications and talks index is being expanded. Current public activity and technical updates are available through the channels below.
Experience
Research Scientist, Lexsi AI
April 2024 - Present
Shipped DLBacktrace v2 for explainability across tabular, vision, NLP, and LLM workloads; expanded transformer and MoE support, and improved inference efficiency through relevance quantization and optimized tracing pipelines.
Research Assistant (NLP Lab), CVPR Unit, Indian Statistical Institute
September 2021 - February 2024
Contributed to TrustED for trustworthiness evaluation in deep learning; developed explainable Indic hate speech detection and worked on an IC50 prediction web application integrating deep learning models.
Education
Master of Technology, Indian Statistical Institute
2019 - 2021
Percentage: 77.45
Bachelor of Technology, Bundelkhand Institute of Engineering & Technology
2015 - 2019
Percentage: 71.82
Let's connect
For AI engineering roles, collaboration, or technical discussion, reach out by email or LinkedIn.