Curriculum Vitae
JESSENTH
EBENEZER
MS Computer Science · New York University
LinkedIn /in/jessenth GitHub sm1else-bot jessenth.com
Education
MS Computer Science
NYU Tandon School of Engineering · Brooklyn, NY
Focus: Deep Learning, Human-Computer Interaction, Extended Reality, Artificial Intelligence
Expected Graduation — June 2027
08/2025 — Present
Brooklyn, NY
B.Tech Computer Science and Engineering
VIT University, Chennai
Coursework: Data Structures, Operating Systems, Computer Architecture, Database Management, Applied Linear Algebra, Parallel Computing
06/2019 — 05/2023
Chennai, IN
Skills
Machine Learning & AI
TensorFlow, Keras, PyTorch, Hugging Face Transformers, LLMs, RAG, Scikit-learn, MLflow, AutoML, SHAP, Optuna, Polars, Pandas, NumPy, PySpark, ETL, Databricks, OpenCV, CUDA, TensorRT, Nvidia NeMo
Programming Languages
Python, C++, SQL, JavaScript (Node.js, Express, React), C# (Unity), HTML/CSS
Development Tools
Linux, Shell Scripting, Kubernetes, Docker, Git, CI/CD (GitHub Actions), Azure (EventHub, Data Lake), ClickHouse, Databricks, REST, FastAPI, Postman
Professional Experience
TVS Motor Company
Data Scientist I
08/2023 — 02/2025
Bangalore, IN
Created a Polars-based binary parser microservice for global market telemetry, achieving a 5× reduction in parsing cost and cutting CPU/memory usage by 40%.
Engineered and scaled real-time data pipelines handling telemetry from 7,000+ connected vehicles, managing ingestion and processing via Kafka and ClickHouse.
Developed an ML pipeline for EV Range Prediction using real-world trip data, achieving 95% accuracy with XGBoost.
Designed a Driver Behavior Engine utilizing complex telemetry signals (acceleration/braking patterns) and deployed features to 20,000+ riders.
Relevant Projects
REFRACT — Multi-Agent LLM Code Intelligence Platform
Multi-Agent Orchestration · FastAPI · SSE Streaming · Provider Agnostic
2025
Provider-agnostic LLM orchestration platform routing source code through a sequential pipeline of specialised AI agents — syntax correction, performance optimisation, test generation and documentation.
MoE-inspired routing architecture with real-time SSE streaming output via FastAPI.
MedVisor — Volumetric MRI Rendering in VR/AR for Surgical Planning
Unity XR Toolkit · DICOM · Raymarching · Volume Rendering · Surface Mesh · Patent Pending
12/2022 — 05/2023
Implemented raymarching, volume rendering and STL-based surface rendering for high-fidelity medical MRI visualization in immersive VR/AR environments.
Optimised Unity XR demo for Quest 2+ hardware, achieving a sustained >90 FPS while processing 2GB+ MRI datasets.
Low-Field MRI Super-Resolution
Medical Imaging · Deep Learning · NYU Neuroinformatics · Beat SOTA Diffusion Baseline
2026
Ensemble model for 4× upsampling of low-field MRI scans, outperforming the state-of-the-art diffusion model baseline.
Topped the NYU Neuroinformatics Spring 2026 Kaggle leaderboard among 36 participants.
SLUICE — Streaming Log Unified Ingestion C++ Engine
C++ · Pybind11 · High-Throughput Log Parsing · Performance Engineering
2025
C++ data parsing engine exposed to Python via Pybind11, built to accelerate ingestion of large telemetry log files.
Achieved 8.25× acceleration in parsing throughput with significant reduction in CPU utilisation.
Forecasting Credit Card Spend for Personalised Customer Relationship Management
Competition Runner Up · Ensemble · 34K Customers
08/2023
Predicted next-month credit card spend for 34,000 customers using 3 months of transaction history and demographic features.
Engineered behavioral spending features and evaluated 14 regression models, building an ensemble prediction pipeline.
Ranked 2nd out of 45 teams with RMSLE = 1.16.
ERICA — GraphRAG AI Tutoring System
GraphRAG · Qwen2.5 7B · MongoDB · NetworkX · Flask
2025
GraphRAG-powered AI tutor representing course material as a knowledge graph for relationship-aware Q&A — understanding how concepts connect rather than just retrieving chunks.
Built on Qwen2.5 7B (local), MongoDB, NetworkX and Flask with a custom frontend.
Publications
Learning Analytics: Virtual Reality for Programming Course in Higher Education
Elsevier · Procedia Computer Science · WEEF 2019 · 51+ Citations
01/2020
Data-driven VR gamification of programming education for college students.
Utilises the Unity3D Engine with XR Toolkit for interaction and MALAR Rubrics for learning analytics.