I build interpretable multimodal AI systems spanning Vision–Language Models · Small Language Models · Clinical AI · NLP · AI Agents
My work focuses on making AI systems evidence-grounded, interpretable, and reproducible, with applications in healthcare, legal intelligence, and public safety.
Website · Research · ORCID · LinkedIn · Hugging Face · CV
How can multimodal AI reason about the world while making its evidence visible?
I explore this question through:
Vision–Language Models Visual understanding, multimodal generation, medical imaging
Explainable AI Evidence attribution, interpretable predictions, model transparency
Small & Efficient Models Small-language-model-guided reasoning and resource-efficient AI
AI Systems RAG, agentic workflows, clinical NLP, legal intelligence
Explainable SLM-Guided Vision–Language Model for Multi-Class Skin Lesion Recognition
Springer Nature · ICICC 2026 · Lecture Notes in Networks and Systems
An explainable multimodal framework combining visual representations with small-language-model guidance for medical image classification.
Top 15% of submissions
India AI Impact Summit 2026 · UN Women × MeitY
A gender-responsive AI initiative focused on women's safety and public-interest technology.
1 of 23 global case studies · 235+ submissions · 50+ countries
An agentic system for clinical coding and revenue integrity that reconciles documentation with ICD-10/CPT codes, identifies discrepancies, and traces findings to supporting evidence.
Virtusa Jatayu Hackathon 2026
Open-source infrastructure for structured, programmatic access to the Constitution of India, designed for legal NLP, retrieval, RAG, and constitutional informatics.
Total PyPI Downloads
| Project | Focus | |
|---|---|---|
| 01 | IndianConstitution | Constitutional NLP |
| 02 | IndianConstitution-js | Legal data infrastructure |
| 03 | SLM-VLM-SkinLesion | Medical Vision–Language AI |
| 04 | CodePerfect Auditor | Agentic clinical AI |
| 05 | Procurement-AI-Agent | AI agents |
| 06 | Visual Quality Inspector | Computer Vision · Quality Intelligence |
Explainable SLM-Guided Vision–Language Model for Multi-Class Skin Lesion Recognition Springer Nature · LNNS · ICICC 2026
NariRaksha: Gender-Responsive AI for Women's Safety India AI Impact Summit 2026 · UN Women × MeitY
Research in progress
IndianConstitution · Architectural Conformance Drift in Autonomous Coding Agents · Explainable Vision–Language Generative Models for Radiology
First Prize ECE Project Contest 2026
Top 15% ICICC 2026 · Springer Nature
1 / 23 Global Case Studies India AI Impact Summit 2026
Invited Reviewer SocialAgent @ NeurIPS 2026
Academic Advisor Machine Learning Tech Society · Saveetha Engineering College
Python · PyTorch · Transformers · Hugging Face · scikit-learn
Vision–Language Models · Small Language Models · LLMs · RAG · NLP · Explainability · AI Agents
FastAPI · Streamlit · LangChain · LangGraph · Git · GitHub · PyPI
Building toward AI systems that combine visual understanding, language generation, evidence grounding, and interpretable reasoning for clinical applications.
Open to research collaboration, applied AI/ML opportunities, and ambitious technical projects.
Website · Google Scholar · ORCID · LinkedIn · Hugging Face · Vikhram Labs
Research: vikhrams@saveetha.ac.in Personal: vikhramselvacumaran@gmail.com
Vikhram S · Multimodal Intelligence · Explainable AI · Healthcare