class AyushNarayan:
def __init__(self):
self.name = "Ayush Narayan"
self.pronouns = "He/Him"
self.location = "Bengaluru, India"
self.education = "BTech ISE @ BMSCE (2023–2027)"
self.research = ["LLM + RAG Systems", "NLP Pipelines",
"Real-Time Comms", "Network ML"]
self.current_work = ["FRPS-2025 Research @ BMSCE",
"Collab — Modelicon Infotech LLP"]
self.interests = ["Full-Stack Dev", "Systems Engineering",
"Competitive Programming", "Cloud"]
self.fun_fact = "Made cursor move using phone gravity."
def say_hi(self):
print("Ship fast. Learn faster. Build things that work.") |
|
| Achievement | Details | |
|---|---|---|
| 🔬 | Funded Researcher | VTU FRPS-2025 · Rs. 4.5L · Modelicon Infotech LLP |
| ☁️ | Google Cloud Certified | 10+ courses — Kubernetes, Terraform, Load Balancing, Scaling |
| 🧩 | Competitive Programmer | 120+ LeetCode — Arrays, Trees, Graphs, DP |
| 🚀 | Hackathon Regular | Winner @ Phase Shift 2024 · Devfolio · Unstop · HackerEarth · MLH |
| 🌐 | Web Coordinator | Info & Web Team — Phase Shift 2024, BMSCE |
🎓 All courses completed & passed on Google Skills Boost
🤖 AI-Powered Academic ERP Assistant — FRPS-2025 Funded, Ongoing
Conversational AI grounding an LLM with live institutional ERP data.
- RAG pipeline reducing hallucination by anchoring on structured ERP records
- Secure API integration between LLM assistant and campus ERP system
- Modular backend for query parsing, auth handling, and ERP request routing
- Targeting campus-wide deployment, publication, and patent filing
Stack: Python LLM APIs RAG REST APIs Modelicon Infotech LLP
🖱️ Mouse Controller — Real-Time Motion Mouse · 🌐 Live Demo
Control your laptop cursor by tilting your phone — like pointing a TV remote at your screen.
- WebSocket-based real-time communication with ultra-low latency
- DeviceOrientation API capturing gyroscope Alpha/Beta/Gamma axes
- Python + FastAPI + pyautogui translating motion → cursor movement
- QR-code pairing, HTTPS/WSS security, live telemetry dashboard
Stack: Next.js React Python FastAPI WebSockets pyautogui Tailwind CSS
📄 Automated Resume Analyzer — NLP Classification Pipeline
End-to-end supervised NLP pipeline for binary resume classification.
- TF-IDF vectorization (max_features=5000) capturing high-signal lexical features
- Gensim tokenization + NLTK stopword removal + L2-regularized Logistic Regression & LinearSVC
- Accuracy, Precision, Recall, F1, confusion matrices; real-time inference workflow
Stack: Python scikit-learn TF-IDF NLTK Gensim Logistic Regression LinearSVC
🌐 Congestion Detection System — Network Systems & ML
Real-time network performance monitoring with ML-based congestion state prediction.
- ICMP / UDP / TCP RTT measurement pipeline + iPerf3 throughput analysis
- Feature engineering: mean RTT, std deviation, signal elevation
- SVM-based congestion classifier; CSV logging + matplotlib graph output
Stack: Python scikit-learn SVM iPerf3 matplotlib NumPy pandas
🇮🇳 India Explorer — Full-Stack Travel Platform · 🌐 Live Demo
A modern full-stack travel exploration platform showcasing India's cultural diversity, heritage, architecture, and tourism ecosystem through a visually immersive interface.
- 🗺️ Interactive state exploration with dynamic routing and categorized tourism filtering
- 🔍 Real-time intelligent search across states and destinations
- ⚡ FastAPI REST endpoints —
/api/states,/api/search,/api/states/{slug},/api/categories - 🗃️ MongoDB Atlas async document architecture with Motor/PyMongo integration
- 🎨 Framer Motion animations, Tailwind CSS responsive layouts, Axios API communication
- ☁️ Full cloud deployment: Vercel (frontend) + Render (backend) + MongoDB Atlas (DB)
System Architecture:
React Frontend (Vercel) → FastAPI Backend (Render) → MongoDB Atlas
Stack: React.js FastAPI MongoDB Atlas Tailwind CSS Framer Motion Axios Uvicorn Pydantic Vercel Render
Capabilities: TF-IDF Pipelines · RAG Architecture · Hallucination Mitigation · NLP Preprocessing · Supervised Classification · Model Evaluation · Confusion Matrices
- DSA: 120+ LeetCode — Arrays, Linked Lists, Trees, Graphs, DP; O(n²) → O(n log n) optimizations
- Core: OOP · System Design · DBMS · OS · Computer Networks · Distributed Systems
- Data: NumPy · Pandas · Matplotlib · TF-IDF · Confusion matrices
- Networks: RTT profiling · ICMP/UDP/TCP · Throughput analysis · Congestion modeling
