Skip to content
available for work remotely✦127.0.0.1✦B.Tech CSE ’27

Navya Anand

AI Systems & Full-Stack Engineer. I build software, and I get paid to judge AI, lately living in LLM evaluation & RLHF▋

◆ panda.dev● coding
currently
  • →Auditing GenAI task queues as QMA at Outlier AI
  • →Designing a deterministic economy-sim game (early prototype)
  • →Building this workstation, my portfolio as an operating system
  • →Exploring LLVM internals and computer architecture
8
projects shipped
5+
freelance builds
4+
security gigs
70%
manual work cut
Python✳TypeScript✳LLM Evaluation✳RLHF✳FastAPI✳React✳Next.js✳PyTorch✳Apache Spark✳PostgreSQL✳Docker✳Prompt Engineering✳Node.js✳Security✳OpenCV✳Linux✳Python✳TypeScript✳LLM Evaluation✳RLHF✳FastAPI✳React✳Next.js✳PyTorch✳Apache Spark✳PostgreSQL✳Docker✳Prompt Engineering✳Node.js✳Security✳OpenCV✳Linux✳
00▸~/terminal

Talk to the machine

This whole site is driven like an OS. Here's a real terminal. Try `help`, `projects`, `open avca`, or `neofetch`. Same engine as the ⌘K palette and the ` shortcut.

navya@workstation: ~
navya@workstation
─────────────────
role AI Systems & Full-Stack Engineer
os Arch Linux (btw) / Hyprland
shell zsh + a terminal for everything
editor Neovim (Lua-configured)
degree B.Tech CSE, Bennett University '27
languages Python · TypeScript · C++ · SQL · Bash · Lua
specialty LLM evaluation · RLHF · full-stack product
host 192.168.1.1
uptime since 2005
8 projects · 7 roles
Interactive shell. Type `help` to begin, `exit` to close.
navya@workstation:~$
▋
try these
anywhere:⌘Kpalette ·`terminal
01▸~/projects --featured

Selected work

Things I've actually shipped, security tooling, generative AI, computer vision, and full-stack products. Each one opens into the real engineering story: the problem, the architecture, the tradeoffs.

All projects

NetSec Analyzer

~/projects/netsec

DNS-tunneling detection over large-scale packet data using Apache Spark, with an automated analytical report.

shippedPythonApache SparkJupyter

Advanced Voice Conversion App

~/projects/avca

Voice-identity transformation using Wav2Vec2 and neural encoders, turn one speaker's voice into another while keeping the words.

shippedPythonPyTorchWav2Vec2

R-CNN Object Detection Studio

~/projects/rcnn

R-CNN object detection with bounding-box visualization and an interactive Streamlit interface for inspecting predictions, applied to marine-debris detection.

shippedPythonPyTorchOpenCV

Lung Cancer Classification (Deep CNNs)

~/projects/onco

Trained and fine-tuned five CNN architectures (VGG19, MobileNet, ResNet, EfficientNet, Inception) for lung-cancer image classification.

shippedPythonTensorFlowKeras
02▸~/capabilities

What I reach for

Honest proficiency, the tools I actually own in production, not a wall of logos I once npm-installed.

Full matrix

AI & LLM Systems

sys/ai
LLM EvaluationRLHF WorkflowsPrompt EngineeringAI Code ReviewRoot Cause AnalysisConversational AI

Machine Learning

sys/ml
PyTorchTensorFlow / KerasCNNs & Transfer LearningOpenCVWav2Vec2 / Audio MLModel Training & Evaluation

Languages

sys/lang
PythonTypeScript / JavaScriptC++SQLBashLua

Backend & Systems

sys/backend
REST API DesignFastAPINode.js / ExpressDjangoJWT Auth & RBACData Modeling

Frontend & Product

sys/web
ReactNext.jsTailwind CSSMotion / AnimationEnd-to-End Product DevFigma

DevOps & Data

sys/ops
DockerLinuxGit / GitHub ActionsApache SparkPower BIPandas / Matplotlib / Plotly
03▸~/mission-log

Where I've worked

From my first engineering job at 17 to auditing frontier-model training data today. Six years, no gaps.

Mission log

Got something you want to build?

I'm a student who ships real things fast, and I'm genuinely curious about hard problems. If you're hiring, collaborating, or just want to talk shop, my inbox is open.

Open a channel Connect on LinkedIn