Python Developer • AI/ML Engineer • Jaipur, Rajasthan

Tanu Meena

"Engineering resilient agentic workflows and high-precision computer vision architectures with uncompromising technical rigor."

Graduating in Artificial Intelligence and Data Science from MBM University (CGPA 7.61), I bridge deterministic software orchestration and advanced neural computation. My core expertise spans stateful multi-node workflows using LangGraph and LangChain, model-driven retail analytics with YOLO and ResNet18 at Tata Consumer Products, and clinical risk evaluation models for DRDO.

Dedicated to architecting production-ready systems that optimize resource efficiency, enforce strict state boundaries, and deliver reliable decision-making intelligence at scale.

Explore Codebases
Tanu Meena, AI/ML Engineer
Core Domains

Specialized Technical Verticals

SOSEO TECH Internship

Agentic Orchestration & LangGraph

Designing stateful LLM workflows using LangGraph and LangChain. Specialized in sequential, conditional, looping, and multi-input task execution with Model Context Protocol (MCP) for secure, tool-connected reasoning engines.

State Management
Tool Calling
Cyclic Flows
MCP Concepts
Tata Consumer Products

Computer Vision Pipeline

Two-stage detection and similarity pipeline using YOLO and ResNet18 for SKU matching and planogram analysis, with documented evaluation limits and failure modes.

DRDO

Clinical ML

ROC-AUC optimized heart failure risk prediction models using structured clinical datasets.

Full Stack

Deployment & APIs

FastAPI & Flask model serialization, OpenCV preprocessing, and Git workflows.

Selected Work

Technical Deployments & Architecture

Explore System Case Studies & GitHub Repositories

01

LangGraph Agent Workflows

Stateful multi-step AI task execution

Problem

Complex LLM applications require deterministic state management, cyclic execution, and secure tool orchestration over multi-step reasoning.

Approach

Engineered graph-based reasoning workflows employing sequential, conditional, and looping logic with explicit state boundaries to guarantee reliable context handoffs.

Technology Stack

Python LangGraph LangChain MCP
GitHub Profile

Interactive Architecture

Select nodes to inspect pipeline
State Initialization

Defines the explicit typing and schema for the graph payload, ensuring strict data boundaries before multi-step execution begins.

02

Shelf Product Identifier

Automated SKU planogram compliance

Problem

Manual retail shelf analysis is inefficient and error-prone, requiring an automated system capable of recognizing diverse SKUs from unstructured images.

Approach

Designed a robust two-stage pipeline separating product localization (YOLO) from specific category classification (ResNet18) to improve overall precision.

Result

85 detections

Saved sample run

Technology Stack

PyTorch YOLO ResNet18
GitHub Repo

Interactive Architecture

Select nodes to inspect pipeline
Image Ingestion

Utilizes OpenCV and NumPy arrays to structurally load and preprocess unformatted retail shelf imagery.

03

Heart Failure Prediction

Clinical risk evaluation model

Problem

Predicting heart failure risk requires accurate processing and evaluation of structured clinical datasets to assist in medical decision-making.

Approach

Developed a machine learning pipeline evaluated with strict classification metrics and ROC-AUC optimization on clinical data.

Technology Stack

Python Scikit-Learn Data Science
GitHub Profile

Interactive Architecture

Select nodes to inspect pipeline
Clinical Data

Ingestion and structured preprocessing of raw clinical datasets containing patient health indicators.

Chronology

Professional Experience

Jan 2026 — Jul 2026

AI/ML Intern • SOSEO TECH Advisory (Delhi)

Developed and tested Python-based AI agent workflows using LangGraph and LangChain for stateful, multi-step task execution. Implemented sequential, conditional, looping, and multi-input workflows with explicit state transitions.

Jun 2025 — Aug 2025

ML Intern • Tata Consumer Products Limited (Bengaluru)

Developed an end-to-end Python computer-vision pipeline using YOLO and ResNet18 for automated retail shelf-image analysis, with evaluation boundaries documented alongside the project.

Jun 2024 — Jul 2024

AI-ML Intern • DRDO (Jodhpur)

Developed a Python machine learning pipeline for heart-failure risk prediction using structured clinical data, evaluated with classification metrics and ROC-AUC.

Beyond the Code

Expeditions & Co-curricular

Travel Expeditions

Karnataka & Tamil Nadu

Coorg, Chikmagalur, Mysuru, Bengaluru, Ooty, Coonoor

Uttarakhand

Rishikesh, Ukhimath, Mussoorie, Rudraprayag, Tungnath, Chandrashila Peak

Rajasthan & NCR

Jaipur, Jodhpur, Udaipur, Delhi

Interests

Hobbies & Versatility

Passionate about trekking, exploring, hiking, and listening to music. Equipped with diverse skills outside of code.

Photography Videography & Editing Social Media Mgmt Farming & Gardening
Community

Volunteer Work

  • Technical Team Volunteer
  • Marketing & Designing Team
  • City Youth Program
Recognition

2021 Award

Awarded a laptop from Super 30 coaching for JEE by IOCL, CSRL (Jaipur).