🤝 Top Collaborators
🤓 Latest Submissions
Vision-Language AI for Medical Image Understanding
MedAgent is a multimodal AI assistant designed to bridge the gap between complex medical imaging and human understanding by transforming raw visual data (X-ray CT scan, medical image) into structured, explainable clinical reports. At its core, the system is powered by the Gemini 2.5 Flash Vision-Language Model (VLM), which allows it to jointly reason over both images and natural language text. Developed using Python, Gradio, and Google Colab, MedAgent provides an accessible browser-based interface where users can upload various medical scan such as X-rays, MRIs, and CT scans and pose natural language questions about them. The system's architecture follows a structured agentic loop that includes an AI critique and self-review process to ensure high-quality output. Each analysis results in a comprehensive report containing: Structured Observations: Precise clinical descriptions of visible features. Medical Findings: Interpretations aligned with standard radiology conventions. Plain-Language Explanations: Accessible summaries tailored for students and non-experts. Uncertainty Notes: Transparent flagging of ambiguous or low-confidence areas. While MedAgent is explicitly not a diagnostic tool and does not replace a licensed physician, it serves as a powerful resource for medical education, interactive research analysis, and preliminary triage support. By combining advanced multimodal reasoning with transparent uncertainty flagging, MedAgent enables users to explore imaging cases with guided, AI-assisted insights.
19 May 2026
Multi-Agent Clinical Reasoning Framework
# Multi-Agent Clinical Reasoning System The Multi-Agent Clinical Reasoning System is a lightweight collaborative AI framework designed to evaluate how iterative multi-agent reasoning workflows can improve clinical decision-making, diagnostic consistency, and reasoning transparency in healthcare-focused AI systems. Rather than relying on a single language model response, the project introduces a structured reasoning pipeline where multiple specialized AI agents collaborate to analyze, critique, and refine medical reasoning outputs. The system is built around three coordinated agents: * **Solver Agent** : Generates an initial clinical interpretation and diagnostic reasoning pathway. * **Critic Agent** : Reviews the generated reasoning to identify logical inconsistencies, weak assumptions, or missing evidence. * **Refiner Agent** : Produces an improved and more reliable final response by incorporating critique feedback into the reasoning process. This workflow simulates iterative expert review commonly used in real-world medical decision-making environments, where multiple specialists contribute to refining diagnostic conclusions. The project uses the MedQA benchmark dataset to evaluate reasoning performance on healthcare-related question-answering tasks. Lightweight open-source language models are used to ensure efficient inference and accessibility in cloud-based environments such as Google Colab. ### Technologies Used * Python * CrewAI * Transformers * PyTorch * Gradio * Qwen2.5-1.5B-Instruct * TinyLlama-1.1B-Chat-v1.0 * Google Colab * Hugging Face Datasets The system architecture is modular and extensible, allowing additional reasoning agents, evaluation layers, or domain-specific healthcare workflows to be integrated in future iterations. The project demonstrates how coordinated multi-agent AI systems can support more transparent, reliable, and scalable reasoning pipelines for healthcare AI research and intelligent decision-support applications.
10 May 2026
TruthLens AI
Every day, millions of people consume breaking news, Instagram reels,Tik Tok videos, tweets without realizing how much of it is AI-generated, manipulated, or misleading. As synthetic media becomes indistinguishable from reality, users lose the ability to verify what’s true. No universal verification layer exists—and trust is collapsing. TRUTHLENS AI is a next-generation Opus-powered workflow that restores transparency to digital content. It ingests multi-format sources (RSS feeds, Reddit posts, mock social media inputs) and analyzes each item through a chain of specialized AI agents that detect: • AI-generated or low-quality “AI slop” content • Manipulated or synthetic images • Misinformation and risky narratives • Suspicious linguistic or behavioral patterns • Factual claims requiring validation The system blends deterministic rules (required fields, thresholds, matching, formatting checks) with advanced LLM reasoning for categorization, summarization, and authenticity scoring. All processing is traceable and auditable, producing a step-by-step provenance record for every item. High-impact, ambiguous, or low-confidence cases automatically escalate to an Agentic Review Layer, followed by Human-in-the-Loop review for final validation—ideal for newsroom teams, policy analysts, or trust & safety units. Outputs include: • authenticity verdicts • risk labels • confidence scores • extracted evidence • external reference links • timestamps, IDs, and full audit summaries Final results are delivered to external destinations (e.g., Google Sheets, email, dashboards) with complete transparency. TRUTHLENS AI demonstrates an industry-ready approach to content authenticity, scalable moderation, and trust verification—fully automated through Opus while remaining human-accountable.
19 Nov 2025
