NATHANI
RAJITHA
Building multi-agent systems, RAG pipelines and applied ML tools, on a foundation of Electronics & Communication Engineering.
ABOUT
INTRODUCTIONWhat if software could actually think?
I'm Nathani Rajitha, an Electronics & Communication Engineering student at VVITU, India. But the real story isn't the degree — it's what happened while I was supposed to be studying.
I built multi-agent systems and real-time RAG pipelines. As an AI Trainee at Lifebox Next Gen, I engineered content automation platforms and intelligence systems that research, reason, and act with minimal human intervention.
For me, engineering has never been about writing code that compiles. It's about closing the gap between what technology can do and what people actually need — building intelligence that feels effortless, not engineered.
That's what I'm here to build.
CURRENTLY IN MY FINAL YEAR — AND JUST GETTING STARTED.“Learning never stops, and neither does growth. As time passes, the builder within evolves across every dimension.”
FUNCTIONAL
EXPERTISE
AI & MACHINE LEARNING🤖
Developing intelligent conversational agents, customized AI co-pilots, and predictive models using Gemini API, LangChain, and advanced NLP.
BACKEND ARCHITECTURE⚙️
Designing scalable backend infrastructures, robust APIs (FastAPI/Node), and automated data ingestion pipelines for real-time analytics.
CLOUD & SYSTEMS🌐
Configuring robust cloud networks, managing databases, and administering systems grounded in Google IT Support and Cloud Computing fundamentals.
Classified Blueprints.
* Handcrafted research files and multi-agent architecture schemas.
Instant Blog Generator
Multi-Agent AIBuilt a full pipeline where a crew of specialized AI agents — Researcher, Writer, Editor, and Publisher — collaborate to produce high-quality blog content on any topic. The system leverages Gemini 1.5 Pro for reasoning and FireCrawl for live web research.
Pipeline Flow:
Technical Highlights:
- Multi-agent collaboration with CrewAI
- Real-time web research with FireCrawl
- Gemini 1.5 Pro for high-quality content
- Automated SEO optimization
- One-click publish to Firebase
Google Screenshot Search Agent
Vision + Agentic AIDeveloped an intelligent agent powered by Gemini Vision that analyzes screenshots pixel-by-pixel, extracts meaningful context, formulates optimal search queries, browses results, and synthesizes a comprehensive answer — all without human intervention.
Pipeline Flow:
Technical Highlights:
- Multimodal vision + text understanding
- Autonomous browser control with Playwright
- LangGraph state machine for agent flow
- Zero human-in-the-loop needed
- Sub-10s response time
Home Security AI System
Computer VisionEngineered a complete home security pipeline using computer vision for person/object detection, integrated with Firebase for real-time alerts. The system runs efficiently on edge devices and provides a live monitoring dashboard with historical incident logs.
Pipeline Flow:
Technical Highlights:
- Real-time YOLO person detection
- Edge-optimized inference pipeline
- Firebase real-time alert system
- Historical incident logging
- Live monitoring dashboard
YouTube Search Project
RAG + Semantic SearchBuilt a semantic search system over YouTube content that goes beyond keyword matching. The system understands user intent, retrieves semantically relevant videos using vector embeddings, and generates concise summaries of video content using transcript analysis.
Pipeline Flow:
Technical Highlights:
- Semantic search with pgvector
- Intent-aware query understanding
- Automatic transcript summarization
- PostgreSQL + pgvector backend
- Sub-second semantic retrieval
Newspaper Wall.
* Taped journal articles and corporate records of applied projects.
AI Trainee @ Lifebox Next Gen
- ✦Building AI-powered applications with Python, FastAPI, LangChain, CrewAI and the Google Gemini API.
- ✦Designing LLM-based multi-agent workflows for content generation, research automation and document Q&A.
- ✦Applying RAG, prompt engineering, vector embeddings and REST APIs for context-aware applications.
- ✦Integrating Tavily Search, DuckDuckGo Search and YouTube API for real-time retrieval.
🔍 key Takeaway:
Multi-agent configurations using Gemini and Tavily Search reduce content hallucination rates by up to 40%.
💡 Stack Implemented:
Python, FastAPI, LangChain, CrewAI, Gemini API, RAG, Playwright.
Hanging Credentials.
NATHANI RAJITHA
B.Tech Student · ECE
B.Tech, Electronics & Communication Engineering
2022 — 2026VVITU · CGPA 8.31
BIEAP — Physics, Chemistry, Mathematics
2020 — 2022Narayana College · 96.4%
SSC
2020Sri Chaitanya School · Grade 10
My Story
TIMELINE
SSC Graduation
Completed Grade 10 secondary education with outstanding academic performance.
Intermediate (BIEAP)
Completed board education in Physics, Chemistry, and Mathematics with a score of 96.4%.
B.Tech Journey Begins
Started B.Tech in Electronics & Communication Engineering, laying the foundation for hardware and computing systems.
Contactless Attendance System
Developed a secure face-recognition-based attendance system using OpenCV, Machine Learning (LBPH), and Tkinter.
Skill Gap Analyzer
Built a Python/Streamlit resume parser using Sentence-Transformers, spaCy, and local models to match profiles with job requirements.
Instant AI Blog Generator
Engineered an AI blog generation platform with CrewAI and Gemini API, coordinating autonomous agents for research, editing, and formatting.
Graduation
Graduated with a Bachelor of Technology in ECE, securing a CGPA of 8.31.
AI Trainee
Joined Lifebox Next Gen, building scalable multi-agent systems, prompt flows, and production-ready RAG pipelines.
Verified Learning.
* Official professional credentials and academic stamps.