Sreenaina Koujala

Full Stack & AI Developer

Currently developing

B.S. Honors in Computer Science | May 2026

AI-Adaptive Learning Platform | Nov 2025 – Present

Generates personalized learning paths from user-uploaded materials via a concept dependency knowledge graph; adaptive assessments dynamically adjust difficulty based on identified knowledge gaps
Rubber Ducky voice companion evaluates mastery via the Feynman technique—students explain concepts aloud, and the AI identifies misconceptions and reinforces weak areas in real time

AI-Powered 3D Digital Twin Platform

Full stack serverless platform that converts natural language building descriptions into interactive 3D models for spatial reasoning, with a xeokit WebGL viewer and AWS orchestration (Lambda, Step Functions, Bedrock)
Implementing MediaPipe hand tracking for gesture-based model manipulation (rotate, zoom, pan), WebXR AR preview for on-site model overlay, and voice-controlled navigation via Whisper for hands-free inspection

AI Dev Workspace | VS Code Extension

Multi-model orchestration engine routing tasks by model strength (GPT-4 for architectural planning, Gemini for research synthesis, Claude for implementation) with persistent project memory and context-aware prompt chaining
Voice-first interface via Whisper + Eleven Labs with self-critiquing architecture recommendations that evaluate trade-offs before suggesting solutions

Maritime & Planetary Intelligence System

Real-time computer vision pipeline (YOLOv8, OpenCV, FastAPI) performing vessel detection, persistent multi-object tracking via Hungarian assignment and IoU matching, and behavioral anomaly detection (loitering, convergence, restricted-zone violations)
Designing planetary-scale extension into wildlife migration tracking and natural disaster monitoring with satellite imagery integration and multi-source sensor fusion

Full-Stack Developer (Part-Time) | May 2018 – Present | Ashburn, VA

Contributed to production data visualization platform deployed across 10+ federal programs, serving government analysts in live decision-making workflows
Rebuilt high-volume data grid with ES6 and Intersection Observer API, enabling stable lazy rendering for 100K+ row datasets and eliminating scroll failures
Engineering AI agents using Claude SDK to automate UI widget generation from natural language specifications, reducing manual frontend development cycles

Full-Stack & AI Co-op (Contract) | Sep 2025 – Present | Fairfax, VA

Led a team to architect and ship a full-stack serverless text-to-3D digital twin platform converting unstructured BIM data into IFC4-compliant models; AWS backend (Lambda, Step Functions, Bedrock, S3, DynamoDB, API Gateway)
Built 8 Lambda functions with a Python IFC generation container, Claude Sonnet for spatial reasoning and entity extraction, and Pydantic-validated scene schema ensuring structural correctness

AI & Data Exploitation Intern | Jun 2025 – Aug 2025 | McLean, VA

Built enterprise chatbot platform using MCP and Strands SDK providing schema-validated, governed access to internal datasets with streaming multi-turn responses
Replaced ad-hoc query workflows with a structured, auditable AI interface—reducing analyst data retrieval time and enforcing access governance

Software Engineering Intern | Jan 2023 | Reston, VA

Migrated React application to functional component architecture for an RF signal processing feature used in spectrum analysis tooling
Built real-time spectrogram visualization with custom scrolling, WebSocket data streaming, and token-based authentication