Projects | Rabimba Karanjai

Bay Area Pujo Parikrama

Tue, 01 Sep 2026 00:00:00 +0000

is a comprehensive community guide and route planner for Durga Puja celebrations across the San Francisco Bay Area.

Key Features:

31 Pujas Across 6 Regions — complete listings covering South Bay, East Bay, Tri-Valley, Central Valley, Peninsula, and Sacramento
Parikrama Route Planner — interactive routing tool to plan pandal-hopping itineraries across multiple celebrations in a single day
Real-Time Ritual Tracking — schedules, bhog availability, ticketing info, and next-ritual countdowns aligned with the panjika calendar
Bilingual Experience — Bengali and English cultural context, including Mahalaya traditions and historical background

Houston Pujo Parikrama

Tue, 01 Sep 2026 00:00:00 +0000

is a community-driven portal and route planner mapping Durga Puja celebrations across Greater Houston, Texas.

Key Features:

Regional Coverage — includes major Houston institutions including Houston Durga Bari Society (Biswa Bangla Sarad Samman winner), SDKKM Temple, Mantra Pearland, and Maitri in Houston
Parikroma Route Planner — route planning across Energy Corridor, West Houston, Southwest Houston / Pearland, and surrounding areas
Full Festival Calendars — schedules spanning Mahalaya, Bodhan, Shashthi–Dashami, and Bijoya Sammilani
Community-First Details — free entry tags, bhog timings, cultural program headliners, and direct navigation links

Dr.MRI.AI

Mon, 18 May 2026 00:00:00 +0000

Dr.MRI.AI is a web-based DICOM viewer built on a simple observation: the hard part of medical AI isn’t the model, it’s knowing what to send it. A typical MRI study contains 200+ slices, but only a handful carry the signal. Dr.MRI.AI plans the slice selection before any multimodal analysis runs.

How it works — Plan → Review → Synthesize:

Plan: Gemma 4 produces a structured plan (which series, slice range, window/level) from your clinical question in plain English. You can accept, edit, or reject it.
Review: The multimodal model analyzes only the focused subset — 200+ slices reduced to ~16 frames, ~99% fewer image tokens.
Synthesize: A final text-only pass produces a structured report with clickable slice references; export findings plus reviewed evidence as one bundle.

Privacy-first architecture:

DICOM files are processed entirely in the browser — no data is uploaded to any server
Fully browser-local inference via onnx-community/gemma-4-E2B-it-ONNX through Transformers.js + WebGPU
Provider-agnostic: Gemini API, OpenAI-compatible endpoints, Ollama (MedGemma planning + Gemma 4 review), or fully on-device Gemma 4 Browser

Built with React 19 + TypeScript + Vite and Cornerstone3D for medical image rendering. Featured in the Google Developer Experts May–Jun 2026 AI community roundup. Educational/research tool — not certified as a medical device.

AI Race Coach (Project Koru)

Fri, 01 May 2026 00:00:00 +0000

The AI Race Coach (Project Koru) is a real-time coaching system for motorsports, developed during the Google I/O GDE Field Test at Sonoma Raceway. The project demonstrated how agentic AI can bridge domain gaps between software engineering and racing expertise.

Architecture:

Antigravity — stateful orchestration and telemetry ingestion
Gemini API — cloud-based driver modeling and strategic insights
Gemma 4 12B — on-device edge intelligence for zero-latency audio coaching
Jetpack Compose — high-refresh Android dashboard for real-time visualization
Google ADK — multi-agent orchestration framework
Custom USB Interface — 10 Hz telemetry stream wired directly from vehicle sensors to Pixel 10

The system identified performance gains beyond human perception — including a new throttle application zone mid-corner in Turn 2 that secured a 0.1-second advantage. Featured on the Google Developers Blog and covered in my blog post Racecraft: The Origin Story.

ConferenceRank

Thu, 01 Jan 2026 00:00:00 +0000

is a unified platform for discovering, evaluating, and tracking computer science conference and journal venues. It aggregates public ranking and bibliometric data into an accessible, interactive interface backed by programmatic APIs.

Key Features

Conference & Journal Rankings — Complete CORE / ICORE rankings (A*, A, B, C) with historical tier tracking, side-by-side with SCImago Journal Rank (SJR) quartiles (Q1–Q4), H-index, and FoR discipline categories.
Venue & Journal Suggester — Abstract-matching engine using TF-IDF and discipline lexicons to recommend stretch, target, and safe venues for an upcoming paper.
Deadlines & Watchlist — Submission countdowns normalized to Anywhere on Earth (AoE), with 1-click Google Calendar sync, .ics export, and zero-login local browser watchlists.
Acceptance Rate Trends — Decades of acceptance statistics (submitted, accepted, rates) for leading venues across systems, AI/ML, security, databases, theory, and graphics.

Built for AI Agents

ConferenceRank is machine-consumable out of the box:

MCP Server (npx conferencerank-mcp) — Exposes six typed tools for MCP-compatible agents: search_venues, get_venue, compare_venues, suggest_venues, upcoming_deadlines, and acceptance_stats.
Static JSON API — CDN-cached endpoints for conferences, journals, and upcoming deadlines, regenerated on weekly automated GitHub Actions runs.
Agent Skill & llms.txt — Ships full instructions and machine-readable context for agentic workflows.

LLM Reasoning & Code Generation

Wed, 01 Jan 2025 00:00:00 +0000

My research on LLMs spans reasoning evaluation, automated code/test generation, and reliability. Key contributions include:

AgentTester — A multi-agent framework for automated unit test generation that significantly improves software reliability (AIWare 2025).
HPCAgentTester — Extending multi-agent test generation to high-performance computing environments.
LogBabylon — A unified framework for cross-log file integration and analysis using LLMs.
Hallucination Mitigation — Techniques for reducing AI hallucinations in diagnostic and clinical contexts.
Evaluating LLM Rationality & Randomness — Benchmarking the quality of reasoning and entropy in LLM-supported tasks.
Synthesizing Public Opinions — Using LLMs to synthesize democratic discourse (IEEE ICEDEG 2025).

This line of research explores how multi-agent systems can produce more reliable, verifiable AI outputs, with applications from software engineering to healthcare.

Quantum Computing & AI

Wed, 01 Jan 2025 00:00:00 +0000

Exploring the intersection of quantum computing and machine learning:

QuCoWE — Quantum Contrastive Embeddings with Variational Circuits for near-term quantum devices (AAAI 2026). This work introduces a novel approach to learning embeddings using quantum variational circuits, designed to work within the constraints of current NISQ-era hardware.
Texas Quantum Initiative — Serving on the Scientific Advisory Committee, helping shape the state’s quantum research strategy.

This emerging research direction investigates how quantum advantages can be practically realized for machine learning tasks, even with today’s noisy intermediate-scale quantum devices.

Blockchain & Decentralized Systems

Mon, 01 Jan 2024 00:00:00 +0000

This research thread represents my largest body of work, spanning smart contract security, decentralized infrastructure, and cross-chain interoperability. Highlights include:

Securing Multi-Chain Ecosystems — A unified agent-based framework for vulnerability detection in Solidity and Move smart contracts (AIWare 2025).
DIaC (Decentralized Infrastructure as Code) — Re-imagining infrastructure management using blockchain (IEEE TNSM 2023).
Smart Contract Translation — Automated translation between smart contract languages using LLMs (FSE 2024).
VerifyGen-X — Cross-chain smart contract generation with verification (NeurIPS 2026, under review).
Ransomware 3.0 — Novel mitigation using proof-of-decryptability smart contracts (IEEE ICBC 2025).
FaaS for dApps — Decentralized Function-as-a-Service for emerging applications (IEEE ICBC 2023).
Conditional Cryptocurrency — Privacy-preserving conditional transactions (IEEE ICBC 2021).

Supported by three consecutive Sui Foundation Academic Research Awards ($25,000 each).

Edge Computing & Trusted Execution

Mon, 01 Jan 2024 00:00:00 +0000

This research area investigates how to bring secure, verifiable computation to resource-constrained edge devices:

Trusted Inference at the Edge — Combining smart contracts with trusted execution environments for verifiable edge inference (IEEE ICBC 2024).
Blockchain + TEE — Lessons learned from applying TEEs in blockchain applications, with implications for future secure systems.
Supporting Heterogeneous Infrastructure — Protecting critical infrastructure across diverse computing environments (AsiaCCS 2023).
AI-Based Personalized Interventions — Deploying AI models for health applications on edge devices (SynBio 2025).

This work bridges the gap between the security guarantees of centralized systems and the autonomy of decentralized, edge-deployed applications.

WebVR & Web Technologies

Mon, 01 Jan 2018 00:00:00 +0000

During my Mozilla Research Fellowship and MS at Rice University, I worked on making virtual reality accessible through web browsers:

Optimizing VR on WebGL — Research on improving parallelism and rendering performance for WebVR applications in Firefox (MS Thesis, Rice University).
Mozilla Research Fellowship — Conducted in-depth research to optimize VR performance in the browser, contributing to Firefox’s WebVR capabilities.
Google Developer Expert — Recognized expertise in Web Technologies, contributing to the broader web platform community.

This work laid the foundation for my continued interest in bringing powerful computing capabilities to resource-constrained environments, whether that’s a browser or an edge device.