🎓 Computer Science @ Ashland University (GPA: 3.8/4.0)
🤖 AI Systems Engineer | 🌐 Full-Stack Architect | ☁️ Cloud & MLOps Engineer
📍 Mansfield, OH

I build production-grade AI systems end-to-end from model training and hardware optimization to distributed cloud backends and scalable frontends.

I don’t just train models. I engineer complete intelligent systems.

• 🧠 Design and fine-tune Transformer & Vision models
• ⚡ Optimize AI for CPU / GPU / TPU / Edge devices
• 🏗 Architect scalable full-stack systems
• ☁️ Deploy secure, cloud-native applications on AWS
• 🔄 Build event-driven microservices & inference APIs
• 📊 Run large-scale benchmarking & bias evaluation studies

Full-Stack Cloud-Native Startup Platform

Impact & Engineering Highlights

• 🏗 Built 77 RESTful API endpoints
• 🔐 Implemented secure JWT rotation + + Role base authentication + OTP verification
• 📦 Designed event-driven email microservice (SQS → Lambda → SES)
• ☁️ Deployed with Docker on AWS EC2
• 📂 Built secure S3 presigned file upload system
• 🧪 Structured logging + production test coverage

Stack: React • TypeScript • Node.js • PostgreSQL • Prisma • AWS • Docker

Real-Time Edge AI Violation Detection

Dual-camera multimodal system detecting 20 traffic violations in real time.

Performance Metrics

• 🎯 99.9% traffic light classification accuracy (LISA dataset)
• 📈 Progressive YOLOv8 training (512 → 640 → 1024 resolution)
• ⚡ FP16 / INT8 optimization with TensorRT
• 🧠 Fusion engine combining spatial + behavioral signals
• 🚀 Edge deployed on NVIDIA Jetson Orin Nano

Stack: PyTorch • YOLOv8 • MediaPipe • ONNX • TensorRT • Docker

LLM-Powered AI Teaching Assistant

Fine-tuned LLaMA 3.2 3B for course-specific reasoning.

Results

• 📈 +70% improvement in academic query accuracy
• 📊 +45% improvement in response quality
• 🔁 Two-phase Supervised Fine-Tuning pipeline
• ☁️ Deployed inference API on AWS EC2
• 🔐 Firebase authentication & feedback loop system

AI Interview Preparation Platform

• 🎥 WebRTC real-time mock interviews
• 📄 Resume parsing + job description alignment
• 📊 Automated scoring & structured feedback
• 📈 40% improvement in candidate performance

AI Multi-Agent Market Intelligence System

• 🤖 Multi-agent orchestration (Product, Competitor, Strategy)
• 🔎 Real-time web data ingestion
• 📊 Automated product research synthesis
• 🧠 LangChain + CrewAI architecture

• 🎥 YOLOv8 + VideoMAE temporal modeling
• 🧠 Multimodal LLM reasoning layer
• 📉 40% reduction in false positives
• 📈 65% improvement in detection accuracy

Natural Language → SQL Desktop Analytics

• 🖥 Electron cross-platform desktop app
• 🔒 Fully local LLM inference (privacy-first)
• 📊 Real-time interactive Plotly dashboards
• ⏱ 80% reduction in manual SQL querying time

Integrated AI + IoT agriculture intelligence system.

• 🌾 30% crop selection accuracy improvement
• 💧 25% water waste reduction
• 🌿 98% plant disease detection accuracy
• ⚡ Benchmarked CPU vs GPU vs TPU performance

• Empirical bias analysis in state-of-the-art Transformer models
• Regression model complexity comparison (IEEE)
• Hardware accelerator performance benchmarking (CPU vs GPU vs TPU)

My research focuses on fairness, scalability, and hardware-aware AI optimization.

I build systems that are:

• Secure by design
• Hardware-aware
• Cloud-scalable
• Edge-deployable
• Measurable and reproducible

I deeply care about performance profiling, structured logging, latency optimization, and production readiness.

• 💼 LinkedIn: https://www.linkedin.com/in/aniketpatel2003
• 🌐 Portfolio: https://aniketpatel.xyz
• 📧 Email: [email protected]