Agentic Systems
Autonomous Discovery Archive
NODE: 2026_RESEARCH_ACTIVEA research archive dedicated to the intellectual pursuit of complex systems, driven by multi-agent Artificial Intelligence orchestration and High-Performance Computing.
Variant Intelligence Cell (VIC)
A multi-agent orchestration framework designed to automate the triage and synthesis of genomic evidence, leveraging Amazon Bedrock AgentCore and Nova Premier with Grounding.
DeepVariant-Inspired Calling
An experimental pipeline with Convolutional Neural Networks for variant calling inspired by DeepVariant, an open-source, deep-learning-based variant caller developed by Google Research.
Agentic Genomic Interpretation
Collaborative agentic systems designed to bridge the gap between human expertise and complex genomic datasets.
Variant Graph Search
Showcases how to empower researchers and bioinformaticians to effortlessly query complex biological knowledge graphs in Neo4j without needing Cypher expertise.
Drug-Therapy-Outcome Search
A modular application to help healthcare professionals and researchers to effortlessly query drug-therapy-outcome relationships without needing specialized database skills.
LoRA Fine-Tuning: FLAN-T5
Efficient parameter-tuning on the SQuAD dataset using Low-Rank Adaptation to optimize model performance for question-answering tasks.
Cyphermind
Dynamic Neo4j Cypher query generation using prompt engineering and KNIME-based LLM integration.
Multi-MCP Algorithm Explainer
A demonstration of a multi-MCP server architecture orchestrated by a Large Language Model for AI algorithm explanations, comparisons, and code examples.
The Investigative Thinking Partner
A multi-agent conversational framework designed to facilitate investigative reasoning and research synthesis in legal contexts.
Research Partner Demo
Recursive reasoning demo designed for high-level research synthesis and investigative logic based on Riskiest Assumption Test (RAT) framework.
Mixed-Model Presentation Deck Assistant
Automated multimodal tool for generating flashcards and content clarity improvements from slide decks.
NumPy vs. CuPy Benchmark
Quantitative performance analysis of CPU vs. NVIDIA GPU-accelerated array processing with CUDA.
Infrastructure Management AI
Developing agentic collaboration protocols for IT infrastructure monitoring and support.
AI-Powered Journal Analyzer
Local-first research analysis utilizing Streamlit and Ollama for secure, on-device data processing.
Golden Moon Food Recommender
Personalized food recommendation engine based on semantic similarity using vector embeddings in ChromaDB powered by a Large Language Model.
Top 5 Destination Scout
LLM-based travel recommendation engine based on cultural significance, adventure, and local sentiment.