Hi, I'm Arnab Maity
AI/ML researcher and software engineer at Johns Hopkins University, working on computer vision, generative AI, and multi-agent systems. Currently enhancing surgical planning pipelines with Video Diffusion Models at the ARCADE Lab.
About Me
I'm Arnab Maity, an AI/ML researcher and software engineer pursuing my Master's in Computer Science at Johns Hopkins University. My journey has taken me from enterprise software at PwC to cutting-edge research — from architecting HIPAA-compliant multi-agent clinical AI platforms at CDHAI to enhancing surgical planning pipelines with Video Diffusion Models at the ARCADE Lab. What started as a fascination with scalable systems has evolved into a passion for building intelligent applications that solve real problems in healthcare and beyond.
I believe in writing clean, thoughtful code and learning from diverse perspectives. I'm particularly drawn to computer vision, generative AI with diffusion models, building multi-agent systems with LangGraph, and creating MLOps pipelines that make intelligent systems accessible and reliable. I'm always curious about emerging technologies but ground that curiosity in practical application – the most elegant solution is often the one that works reliably and makes someone's day easier.
Skills & Technologies
Core languages, AI/ML frameworks, cloud infrastructure, and DevOps tools I use regularly.
Languages
AI/ML & Data Science
Computer Vision & Generative AI
Cloud & Infrastructure
MLOps & DevOps
Databases
Experience & Education
My professional journey and educational background that shaped my development skills
Work Experience
Research Assistant
Johns Hopkins University – ARCADE Lab (Prof. Mathias Unberath)
- •Enhancing BronchOpt, a bronchoscopy trajectory optimization pipeline leveraging Video Diffusion Models with Diffusion Transformers (DiT) and ControlNet conditioning to synthesize anatomically realistic endoscopic video sequences for surgical planning and training data augmentation.
- •Curating and preprocessing paired bronchoscopy imaging datasets with clinical reports; performing data quality validation, consistency checks, and exploratory analysis to ensure reliability for downstream diffusion model training and segmentation pipelines.
Research Assistant
Johns Hopkins University – Center for Digital Health and AI (CDHAI)
- •Architected a HIPAA-compliant multi-agent clinical AI platform using LangGraph with an Orchestrator Agent routing between a Patient Agent (symptom intake, medication clarification, wearable/CGM sensor sync) and a Clinician Agent (EHR pre-analysis, summary generation, anomaly flagging) surfaced via a Clinician Dashboard.
- •Built end-to-end data pipeline streaming Epic EHR data via HL7 into Azure Data Lake with Databricks batch jobs, Unity Catalog for data governance, and MLflow for model registry, with Azure AD SSO, RBAC, and encryption-at-rest for HIPAA compliance.
Associate
PricewaterhouseCoopers (PwC)
- •Led full-stack development for greenfield SAP implementation across 3 modules, building FIORI apps with JavaScript/ABAP and establishing automated CI/CD pipelines, achieving 30% efficiency gains.
- •Architected RESTful API microservices handling 10,000+ requests/hour with automated ETL workflows, improving business processes by 40%; automated infrastructure with Ansible, reducing deployment time by 80%.
Education & Certifications
Masters in Computer Science
Johns Hopkins University
- •Relevant Coursework: Deep Learning, Learning Based Vision, Human-Centered AI, Software System Design
Bachelor’s in Electrical Engineering
Institute of Engineering and Management, Kolkata
- •GPA: 9.04/10
Certifications
Featured Projects
Here are some of the projects I've worked on. Each project represents a unique challenge and learning experience.
PulsePoint
FeaturedAI-powered medical triage system using an adversarial LLM council (GPT-4o, Claude Sonnet, Gemini 2.5 Flash) orchestrated via LangGraph with fast/visual/council routing delivering urgency-classified assessments in under 3 seconds. Engineered RAG with MongoDB Atlas vector search, LiveKit WebRTC for real-time voice I/O, and full observability via Arize Phoenix with OpenTelemetry tracing.
PerceptEye
FeaturedUniversal accessibility platform with real-time sign language detection (ASL to text/speech) and live audio transcription using computer vision models (MediaPipe, PyTorch, Transfer Learning), demonstrating strong CV model implementation and evaluation skills applicable to medical image analysis.
Meridian
FeaturedComprehensive DevOps culture enhancement platform that bridges the gap between learning and doing. Provides AI-powered insights, guided learning journeys, and data-driven management tools to transform how teams approach DevOps.
Hackathons
Selected hackathons I participated in, focusing on Development, Cloud, AI, and rapid prototyping.
NexHacks CMU 2026
Built PulsePoint, an AI-powered medical triage system using an adversarial LLM council (GPT-4o, Claude Sonnet, Gemini 2.5 Flash) orchestrated via LangGraph with fast/visual/council routing, RAG with MongoDB Atlas vector search, LiveKit WebRTC for real-time voice I/O, and full observability via Arize Phoenix.
YCP Hacks 2025
Built PerceptEye, a universal accessibility platform with real-time sign language detection (ASL to text/speech) and live audio transcription using computer vision models (MediaPipe, PyTorch, Transfer Learning).
HackUMBC 2025
Built Meridian, a comprehensive DevOps culture enhancement platform bridging the gap between learning and doing, with AI-powered insights and data-driven management tools.
HopHacks 2025
This project focused on merging empathy and technology — transforming scattered information into easy and meaningful support in moments of need.
Global API Fest 2022
Built a full-stack React.js + Django application, implementing OpenAPI 3.0 standards where it had a collection of APIs like Speech-to-Text, Text-to-Speech, Text-to-Image.
Get In Touch
I'm always open to discussing new opportunities, interesting projects, or just having a chat about technology.
Let's Connect
Whether you have a project in mind, want to collaborate, or just want to say hello, I'd love to hear from you. Feel free to reach out through any of the channels below.
Send a Message
© 2026 Arnab Maity.