What I built: Document layout analysis and OCR experiments for digitizing historical texts.
Software R&D + ML Research
I build ML systems, data pipelines, and research tooling.
I translate research ideas into scalable software with deep learning, multimodal modeling, and distributed systems.
Graduate researcher with open-source ML systems work. Professional background and credentials are listed on LinkedIn.
- Document AI pipeline work for layout + OCR in RenAIssance.
- NLP crisis-signal prototype experiments in ISSR.
- Dataset versioning workflows with MinIO + LakeFS.
Open to: Research Engineer / ML Systems Focus: ML research + distributed systems
GSoC 2025 repository work Public GitHub repos Research notes Robotics demo clips
Research Systems
How I think about ML research, data, and infrastructure.
Three system views that reflect how I scope research problems, build data pipelines, and ship models.
Multimodal ML Pipeline
From data intake to evaluation, with feedback loops for continuous learning.
Distributed Data Platform
Storage, versioning, and compute aligned for reproducible experiments.
Experiment Lifecycle
A research loop optimized for iteration speed and insight capture.
All Projects
Full project timeline
Recent ML research highlights are above. This is the complete catalog across research, systems, and software.
RenAIssance Document Analysis
What I built: Layout analysis + OCR experiments for historical document digitization.
PyTorch LayoutLMv3 Document AI
ISSR — Crisis Detection
What I built: NLP prototype for crisis-signal classification using social text, sentiment, and location features.
NLP Sentiment Analysis Geospatial Signals
MinIO + LakeFS Data Infrastructure
What I built: Object-storage + dataset versioning workflow for reproducible ML experiments.
MinIO LakeFS Dataset Versioning
ArtExtract — Art Analysis AI
What I built: CNN-RNN experiments for artwork classification and style-analysis tasks.
PyTorch CNN-RNN Classification
Additional projects (embodied AI, full-stack, mobile, games)
2025 - ML Research & Systems
PyMyCobot Teleoperation
What I built: Local teleoperation workflow and control scripts for manipulation data-collection experiments using the PyMyCobot stack.
Python Teleoperation Data Collection
UMI Multi-Sensor Data Extraction
What I built: Multi-sensor extraction scripts and trajectory analysis workflow around UMI datasets and ORB-SLAM-based tracking.
ORB-SLAM3 Trajectory Analysis Sensor Fusion
UMI Data Annotation Pipeline
What I built: Annotation workflow for manipulation episodes, including gripper-pose and keyframe labeling around embodied-CoT style experiments.
Dataset Annotation Computer Vision Embodied CoT
MinIO + LakeFS Data Infrastructure
What I built: Object-storage and dataset-versioning setup for robotics datasets with reproducible data snapshots.
MinIO LakeFS Dataset Versioning
LeRobot ISO-101 Platform
What I built: Local data-collection and policy-training workflow using LeRobot tooling for fast experiment loops.
LeRobot ROS2 Policy Training
RenAIssance Document Analysis
What I built: GSoC 2025 repository work on historical document layout analysis and OCR experiments.
PyTorch LayoutLMv3 Document AI
ISSR - Mental Health Crisis Detection
What I built: GSoC candidate repository prototype for crisis-signal detection from social text with sentiment and geospatial features.
NLP Sentiment Analysis Geospatial Signals
ArtExtract - Art Analysis AI
What I built: CNN-RNN project for artwork classification and style-analysis experiments.
CNN-RNN PyTorch Classification
2024 - Full-Stack Development
NutrAI Health Assistant
What I built: Full-stack nutrition app prototype with meal-planning and dietary-analysis features.
React Python Web App
CampusBuddy Mobile App
What I built: Flutter app for campus events, dining, and student-resource discovery.
Flutter Dart Mobile App
Point of Sales Terminal
What I built: Java POS application with inventory tracking, payment flows, and sales reporting.
Java Desktop App Inventory
2023 - Game Development
Space Invader Game
What I built: C++ arcade game project with custom gameplay mechanics and rendering logic.
C++ SFML Game Loop
Recent Writing
Research notes and systems breakdowns
Short technical writeups on robotics, ML systems, and multimodal research.
The State of Robotics in 2025: Why the Hype Isn't Lying (Yet)
A deep dive into the four critical bottlenecks slowing the robotics revolution and why general-purpose robots remain inevitable
Making Sense of Multimodal Models with Partial Information Decomposition
A deep dive into how Partial Information Decomposition (PID) reveals how different modalities interact in AI systems, from redundancy to synergy