Building intelligent systems with Deep Learning, Computer Vision & NLP
class AIEngineer:
def __init__(self):
self.name = "Prathmesh"
self.focus = ["AI/ML", "Computer Vision", "NLP"]
self.stack = ["PyTorch", "FastAPI", "Docker"]
self.mindset = "Product & Scale"
def get_status(self):
return "Ready for Opportunities"
Hey! I'm Prathmesh, a tech enthusiast who thrives on learning by building. My core focus lies in Artificial Intelligence and Machine Learning, where I enjoy taking abstract ideas and engineering them into real, working products. Alongside AI/ML, I work across the full stack — designing clean frontends, building robust backend systems, and deploying them on modern cloud infrastructure.
I approach engineering with a product mindset — not just writing code that works, but building systems that scale, perform, and solve real problems. I'm still early in my journey, but every project teaches me something new, and every bug fixed is a lesson banked.
Specialization in Artificial Intelligence & Machine Learning with focus on Deep Learning, Neural Networks, Computer Vision, and NLP.
Full-stack web application for detecting AI-generated fake images and videos using PyTorch ensemble models, Celery task queues, Redis, and FastAPI containerized with Docker.
Real-time ML face recognition (25 FPS) using TensorFlow MobileNetV2 transfer learning, OpenCV, eye-blink EAR liveness detection, and SQLite attendance logging.
Instant high-priority email notification system via WhatsApp featuring 2-stage AI scoring (0-100), Gmail Pub/Sub & Outlook Graph webhooks, FastAPI, and Celery.
Full-stack community management platform with secure JWT authentication, membership applications, event/meetup management, hackathons, and admin analytics.
E-commerce platform with user authentication, product browsing, shopping cart functionality, and seamless REST API integration with demo credentials.
Cybersecurity tool for hash identification and cracking with support for MD5, SHA-1, SHA-256, and bcrypt featuring CLI and Web UI interfaces.
AI tool that analyzes GitHub repositories using static analysis and GPT-4 intelligence to detect security vulnerabilities and performance bottlenecks.
Voice-enabled conversational AI recruiter conducting technical interviews. Manages audio flow, generating dynamic questions via LLM reasoning.
Modular Python CLI application for high-performance trading on Binance Futures Testnet with order placement and activity logging.
Completed an intensive 3-month AI/ML engineering internship building production ML models with Scikit-learn/TensorFlow, developing FastAPI microservices, and integrating LLM automation pipelines into React web applications.
Advanced certification in Deep Learning with TensorFlow 2 covering custom neural network architectures, CNNs, LSTMs, transfer learning, and model optimization for inference.
Official PCAP-31-03 professional certification demonstrating expertise in Object-Oriented Programming (OOP), module design, exception handling, data structures, and Python best practices.
Designed and deployed a PyTorch ResNet-50 ensemble model achieving 92% accuracy on synthetic media detection benchmarks with real-time video stream anti-spoofing.
Interactive simulations of core deep learning, computer vision, and NLP algorithms.
> Model loaded: ResNet-50 Ensemble
> Status: Frame analyzed cleanly. No manipulation detected.
LLMs, Prompt Engineering, RAG Architectures
Model Deployment, CI/CD for ML Pipelines
Reading latest AI Papers & Architectures
3D Vision, Real-Time Video Stream Processing
Completed 3-month AI/ML engineering internship at Zetpeak
Imperial College London via Coursera
Python Institute / Cisco Networking Academy
Deepfake Detection Model Ensemble Achievement
Do we really need Retrieval-Augmented Generation for every intelligence system? Exploring RAG vs Open Knowledge Format (OKF), fundamental architectural differences, where each wins, and why hybrid routing is the future for AI agents.
I'm always open to discussing AI/ML projects, collaborations, or opportunities.