TIRTH BHANDERI
#opentowork

Hi, I'm Tirth.
Software Engineer.

View Work Resume
Tirth Patel

01 // About

I'm a BTech student from Gujarat with a foundation in software development and AI. I love figuring out how things work under the hood, building intelligent systems, and pushing my engineering boundaries.

Connect

Let's build
something.

languages

JavaPythonDartJavaScriptSQL

frontend

ReactNext.jsTailwind CSS

mobile

FlutterAndroid SDKWebRTCREST integration

backend

Spring CoreSpring BootFastAPIFlaskNode.jsNestJSmicroservices

databases

MySQLPostgreSQLMongoDBFirebaseSQLite

ai ml

LiteRT (TensorFlow Lite)ONNX Runtime MobileMCPRAGLangChainLangGraph

tools

GitGitHubDockerAlembic

SELECTED WORK

Core Architecture

Krushi-Netra (TFLite, Edge AI)

An offline-first Android application that runs real-time object detection on cattle (12 different types of Indian cattles) using TensorFlow Lite. Handled the entire model quantization and Android CameraX integration pipeline to ensure 30fps inference on budget devices without internet access.

Core Architecture🏆 National Level Finalist

Metis (AI Interviewer / LangGraph)

An intelligent mock interviewer agent built on LangGraph and OpenAI. Metis parses candidate resumes to dynamically generate highly contextual technical questions, scoring candidates on both technical depth and communication clarity in real-time.

Core Architecture

Charaka Vaidya (Ayurvedic RAG Assistant)

A specialized Retrieval-Augmented Generation system trained on classical Ayurvedic texts. Employs a complex multi-stage pipeline: intent classification, Sthana-weighted semantic search via ChromaDB, and response synthesis in English, Hindi, and Gujarati.

Core Architecture

ML Model Inference Serving Platform

Deploys ONNX/TFLite models as REST inference endpoints with schema introspection, a versioned model registry, a Redis-backed job queue, and a latency/error dashboard. Stack: Spring Boot, Redis, FastAPI, ONNX Runtime, React.

Open Source🏆 1 State & 2 National Level Finalist

a2a-firewall

An open-source Agent Runtime Security Fabric & Zero-Trust Governance Mesh designed to inspect, authorize, throttle, sandbox, and cryptographically audit autonomous AI agent fleets and multi-agent systems. Key Capabilities: - Ed25519 Cryptographic Identity & Macaroon Attenuated Delegation - Decision Evidence Envelopes & Deterministic Policy Replay - Agent Memory & RAG Write-Time Firewall

Built withPython
Open Source🏆 Podium Finish

AI_Solution_Builder

From Business Intent to Mounted, Production-Ready Software Systems in Minutes. An enterprise-grade AI system that converts business ideas, BRDs, SOPs, and legacy schemas into fully functional, tenant-isolated software applications alongside traditional architecture blueprints.

Built withPython
Open SourcePersonal Project

AegisPay

A secure, Python-based payment processing simulation tool. Implements simulated banking endpoints, fraud detection heuristics, and comprehensive unit tests for high-reliability fintech integrations.

Built withPython

Experience

DEZAI

Backend Engineering Intern

June 2026 - Present

Developing highly scalable AI infrastructure. Designing async microservices with FastAPI and Spring Boot to serve high-throughput embedding generation and retrieval pipelines.

HyTech Education

AI Full-Stack Engineer

May 2026 - June 2026

Appointed directly by the founder to build an AI-powered curriculum compliance tool. Architected the entire solution from database schema to LLM prompt engineering in under 4 weeks.

9Brainz

Android Developer Intern

Dec 2024 - Apr 2025

Developed and maintained highly scalable e-commerce Android applications. Implemented WebRTC for real-time customer support features and optimized local SQLite caching for offline mode.

Education

CHARUSAT

B.Tech in Artificial Intelligence & Machine Learning

Expected 2028

Current GPA: 8.23 (till 2nd Year). Focusing on advanced machine learning algorithms and software architecture.

RK University

Diploma in Computer Engineering

Sep 2023 - Apr 2025

Graduated with 9.63 GPA. Built foundational knowledge in data structures, hardware integration, and algorithms.

Certifications

Verified credentials and professional training.