Divya Natekar
About Research Experience Projects Skills Contact
NYU CUSP · Urban Data Science

Building Scalable AI Software, and Spatial Systems

I am Divya Natekar, an MS Urban Data Science student at NYU CUSP working across agent-based simulation, geospatial analytics, 3D LiDAR point clouds, machine learning workflows, and cloud application engineering.

View Resume Explore Projects Email Me
Brooklyn / New York Urban AI Spatial Data Science Software Engineering
About

Bridging software engineering, spatial data, and AI.

My work combines data engineering and machine learning with complex modeling questions, building software engines that are rigorous, scalable, and decision-ready.

I am pursuing an MS in Urban Data Science at NYU’s Center for Urban Science + Progress. My background spans AI microservices, cloud backend development, agent-based mobility simulations, and computer vision workflows.

I have engineered full-stack Next.js/Supabase AI applications, built 3D LiDAR point cloud filtering pipelines, simulated multi-agent urban mobility decisions in Brooklyn, and analyzed wildfire risks using satellite datasets.

I am especially interested in agent-based simulation systems, generative AI integrations, software engineering for data platforms, and applied spatial machine learning.

01

Agent-Based Modeling

Urban mobility simulations, choice decision engines, spatial interaction modeling, and LLM-assisted context pipelines.

02

Geospatial AI & Remote Sensing

MODIS/FIRMS satellite processing, wildfire risk modeling, spatial statistics, and QGIS visualization workflows.

03

3D Point Clouds & Vision

LiDAR point-cloud processing, RANSAC filtering algorithms, transient object removal, and geometric 3D urban models.

Research

Applied research & simulation models.

Grounded in real-world urban data, sponsor requirements, and advanced computational methods.

Sept 2025 – May 2026

Thinking on the Move: Agent-Based Urban Simulation & Decision Modeling

NYU CUSP · Downtown Brooklyn Partnership

Engineered an agent-based urban simulation framework analyzing lunchtime mobility patterns and multi-agent choice behavior across dense urban corridors in Downtown Brooklyn.

  • Developed agent-based logic simulating individual decision preferences, spatial routing, and corridor-level movement.
  • Integrated gravity-based spatial interaction methods with LLM context utilities to capture qualitative decision drivers beyond simple distance metrics.
  • Processed high-dimensional mobility records, POI datasets, and review embeddings to deliver predictive urban policy evaluations for stakeholders.
Jun 2025 – Aug 2025

Transient Object Removal from Urban LiDAR Point Clouds

NYU CUSP Summer Guided Research · Mentor: Prof. Debra Laefer

Designed point-cloud filtering workflows to remove transient objects such as pedestrians and moving vehicles from high-density urban LiDAR scans.

  • Applied RANSAC plane fitting and spatial clustering algorithms to isolate static urban geometry from moving clutter.
  • Awarded the NYU Tandon CUSP Experiential Learning Scholarship for Summer 2025 guided research.
  • Presented findings at the BATWorks Climate Event and the NYU CUSP Fall 2025 Research Showcase.
2026

Urban Wildfire Risk Prediction using Remote Sensing and Deep Learning

NYU CUSP · Urban AI Project

Engineered an end-to-end remote sensing workflow to predict and visualize wildfire risk using MODIS thermal satellite imagery and NASA FIRMS active fire records.

  • Built automated Python and Google Colab pipelines to extract, clean, and align multi-spectral spatial imagery.
  • Trained PyTorch predictive models using parameter-efficient fine-tuning (PEFT/LoRA) for high-accuracy spatial risk mapping.
Experience

Software engineering & industry impact.

Hands-on experience developing microservices, AI backend pipelines, and real-time cloud systems.

Software Engineer Intern — AI/ML, Lead Engineer

PM Accelerator · Remote · Aug 2026 – Oct 2026 (Expected)

Serving as Lead Engineer to construct backend micro-routes, telemetry monitoring pipelines, and data context infrastructure for an AI-driven personal relationship management product.

  • Architected Next.js App Router API micro-services (/api/people, /api/decisions) backed by Supabase and PostgreSQL.
  • Integrated Langfuse SDK telemetry across API boundaries to log real-time inference inputs, trace prompt execution flows, and optimize latency.
  • Built snapshot transformation utilities mapping asynchronous PostgreSQL state into structured dynamic payloads for prompt evaluations.
  • Refactored query performance using Supabase ISO timestamp boundaries (.gte/.lte), eliminating inefficient regex scans.
  • Led engineering reviews, resolved Git branch merge conflicts, and standardized dev team practices.

Software Development Intern

L&T Technology Services (LTTS) · Airoli, India · Jan 2023 – Jul 2023

Engineered real-time data integration components for a healthcare proof-of-concept connecting wearable devices to the Philips HSDP cloud.

  • Built asynchronous data ingestion pipelines using Java Spring Boot, Apache Kafka, and Spring Vault.
  • Provisioned cloud infrastructure using Terraform scripts and Microsoft Azure cloud services.
  • Configured Azure DevOps CI/CD pipelines, executed API testing with Postman, and documented services using Swagger.
Projects

Selected engineering projects.

A collection of simulation models, remote sensing tools, point cloud processing, and product applications.

Urban Agent Mobility & Choice Engine

Built simulation logic using agent-based modeling, multinomial logit formulations, and spatial interaction techniques to evaluate spatial mobility flows and urban decision behavior.

Agent-Based ModelingPythonSpatial AIMobility Analytics

Satellite Wildfire Risk Pipeline

Created an end-to-end computer vision and machine learning model processing MODIS and FIRMS satellite records to forecast regional fire hazards.

MODIS / FIRMSRemote SensingPyTorchLoRA / PEFT

3D LiDAR Spatial Filter Engine

Engineered algorithmic point-cloud filtering tools using RANSAC and spatial clustering to remove moving obstacles from urban LiDAR scans.

LiDAR3D Point CloudsRANSACCloudCompare

Mech-On-Wheels

Built a real-time roadside assistance app prototype connecting drivers in breakdown scenarios with nearby mechanics; earned Google Startup Weekend honorable mention.

Product PrototypeFull-Stack UXStartup Weekend
Skills

Technical toolkit.

Languages, frameworks, databases, and spatial platforms used across industry and research.

Programming

PythonTypeScriptJavaC++SQLJavaScriptPandasNumPy

AI & Machine Learning

LLMsPyTorchTensorFlowScikit-learnLangfuseRAGComputer VisionRANSAC

Full-Stack & Cloud

Next.jsReact.jsSpring BootSupabasePostgreSQLAzureKafkaTerraform

Spatial & Geospatial

GIS / QGISSpatial AnalyticsAgent-Based Modeling3D LiDARRemote SensingPoint Clouds
Education & Leadership

Academic background.

New York University, Tandon School of Engineering

MS Urban Data Science · NYU CUSP · GPA 3.61/4.0 · Expected May 2026

Manipal University Jaipur

B.Tech Computer Science & Engineering · Minor in Cloud Computing Applications · 2019–2023

Leadership & Involvement

IEEE Student Branch Senior Coordinator, Women in Engineering Lead, and Head of Programs for Randomize(); Computing Club.

Contact

Let’s connect.

I am open to Software Engineering, AI/ML, Data Science, and Geospatial Analytics roles. Feel free to get in touch.

Download Resume dyn2009@nyu.edu GitHub