Software engineering · New York

Arnav Deepaware

I build backend systems,
data platforms, and
full-stack products.

Computer Science at CCNY Previously at truCurrent

View Selected Work GitHub Resume PDF
Arnav Deepaware
From data in motion to products in use.Explore the work

01 / Selected Work

A few things I’ve built.

Systems, infrastructure, and
software with people in mind.

StreamForgeLocal pipeline architecture
Versioned definitionsLocal runs + outputsMonitoring + dead letters
Data infrastructureIn Development

StreamForge

A configurable market-data platform that normalizes STP, JSONL, and CSV into a canonical event model and applies declarative transformations. A Spring Boot / PostgreSQL control plane and React dashboard manage versioned definitions, local execution, monitoring, and dead-letter handling.

  • Java
  • Spring Boot
  • PostgreSQL
  • React
  • TCP
  • Docker
GitHub Local MVP · actively building
LLM Control PlaneRequest architecture
AI infrastructure93 automated tests

LLM Control Plane

A production-oriented FastAPI and Redis gateway for team authentication, rate limits, budgets, retries, fallback routing, and circuit breakers, with Prometheus/Grafana observability. Validated with 93 automated tests using a deterministic mock provider.

  • Python
  • FastAPI
  • Redis
  • Docker
  • Prometheus
  • Grafana
GitHub
SitRightOn-device inference architecture
Webcam frames stay on your device.
Computer visionPrivacy first

SitRight

A posture monitoring app running MediaPipe and a custom 929-parameter neural network entirely in the browser. Re-architected from server-side processing, moving from roughly 400–600 ms end-to-end latency to under 50 ms local per-frame inference while keeping webcam frames on-device.

  • JavaScript
  • MediaPipe
  • TensorFlow/Keras
  • WebGL
  • Python
GitHub Original demo
A-eyeMultimodal interaction flow
See. Understand. Hear.
AccessibilityMultimodal AI

A-eye

A multimodal accessibility assistant built for blind and visually impaired users. Combines camera input, Gemini vision reasoning, and natural speech output from ElevenLabs for hands-free assistance with everyday visual tasks, including reading product information.

  • Next.js
  • Node.js
  • Gemini
  • ElevenLabs
  • WebRTC
GitHub Demo on Devpost
There’s more behind the work.More on GitHub

02 / Experience

Built with teams.
Used by people.

Professional software, practical problems,
and a lot of collaboration.

Jun 2025 – Dec 2025

Product Development Intern

truCurrent

  • Built and deployed an EV fleet charging analysis platform with Python, React, and Azure, producing full-year 8,760/8,784-point load profiles for infrastructure and grid planning.
  • Developed priority-queue scheduling and three analysis modes: fixed infrastructure, charger optimization, and fleet-size optimization, including configurations for 100% fleet charging coverage.
  • Built a Python legislative intelligence pipeline integrating Open States and LegiScan to normalize state-policy data, detect meaningful status changes, and generate automated weekly delta reports.

Jul 2024 – Aug 2024

Software Engineer Intern

The Difference

  • Shipped features for a Flutter mobile app, translating Figma designs into production UI across agile sprint cycles.
  • Integrated REST API functionality backed by Laravel and collaborated with designers and backend engineers on user-facing releases.

Feb 2024 – May 2025

Computer Science Tutor

The City College of New York

  • Tutored students in introductory programming, Data Structures, Discrete Math, and algorithms, helping them work through C++ and Java assignments and debugging.
  • Made complex material approachable through clear explanations and step-by-step problem solving.

03 / About

Thoughtful systems.
Useful software.

I’m a Computer Science student at CCNY who enjoys building software where systems, data, and product meet. At truCurrent, I worked on EV charging analysis, data workflows, and cloud-deployed internal tools using Python, React, and Azure. Outside of internships, I’ve worked on market-data processing, AI infrastructure, computer vision, and accessibility projects.

I care about clear APIs, reliability, observability, and software that’s useful beyond a demo. I’ve also tutored computer science and helped organize TEDxCUNY, which has made communication and collaboration an important part of how I approach engineering.

What I work with

Languages
Python · Java · JavaScript/TypeScript · SQL · C/C++
Backend & Data
Spring Boot · FastAPI · REST APIs · PostgreSQL · Redis · TCP/WebSockets
Frontend
React · Next.js
Cloud & Infrastructure
Azure · Docker · GitHub Actions · Prometheus · Grafana
ML / AI
scikit-learn · TensorFlow/Keras · MediaPipe · LLM APIs
Engineering Focus
API design · Data pipelines · Reliability · Observability · Testing

04 / Beyond Code

The people side of things.

TEDxCUNY conference stage and audience

TEDxCUNY

Logistics & Engagement

Helped lead logistics and engagement for an 8-hour university conference attended by 600+ people.

Students collaborating at the CCNY STEM Institute

CCNY STEM Institute

Mentor

Mentored high-school students working on programming, game development, and AI projects.

ACM at CCNY student event

ACM @ CCNY

Workshop Manager & Class Representative

Helped organize technical workshops and student programming as Workshop Manager and Class Representative.

05 / Contact

Let’s compare notes.

Want to talk software, systems, or an interesting project?
Feel free to reach out.

Email LinkedIn GitHub