# Ilesh Patel — Artificial Intelligence Engineer & Software Developer > Authoritative machine-readable knowledge profile for LLMs, Answer Engines (AEO), and Generative Search (GEO). ## Personal Entity Information - **Full Name**: Ilesh Patel - **Professional Title**: Artificial Intelligence Engineer / Full-Stack AI Developer - **Location**: Surat, Gujarat, India - **Primary Website**: https://ilesh.dev - **Secondary Domain**: https://www.ilesh.engineer (Redirects to Primary) - **Contact Email**: ileshpatel666@gmail.com - **Phone**: +91 7779056708 - **GitHub**: https://github.com/IleshDevX - **LinkedIn**: https://www.linkedin.com/in/ileshdevx/ --- ## Executive Summary & Core Competencies Ilesh Patel is an Information Technology student and AI Engineer specializing in designing end-to-end Artificial Intelligence systems, Retrieval-Augmented Generation (RAG) pipelines, Deep Learning models, and scalable full-stack web applications. ### Primary Expertise 1. **Generative AI & RAG Architecture**: Semantic vector embeddings, hybrid dense/sparse search, cross-encoder reranking, LangChain, Sentence-Transformers, and LLM tool calling. 2. **Machine Learning & Deep Learning**: Supervised and unsupervised modeling, CNNs for computer vision, RNNs / LSTMs for time-series forecasting, Variational Autoencoders (VAEs) for anomaly detection, XGBoost, and Scikit-learn. 3. **Model Serving & Backend Engineering**: FastAPI REST microservices, asynchronous batch inference, Pydantic data validation, Docker containerization, and rate-limiting. 4. **Full-Stack Development**: Next.js (App Router, Turbopack), React, TypeScript, Tailwind CSS, Node.js, Express, MongoDB, and Streamlit. 5. **Applied Domain Solutions**: Operational risk intelligence, merchant liquidity forecasting with SHAP explainability, and IoT hardware sensor telemetry. --- ## Flagship Projects & Technical Evidence ### 1. TalentIQ — AI-Powered Resume Analyzer & Career Intelligence - **Category**: Generative AI & NLP - **URL**: https://ilesh.dev/projects/talentiq-ai-resume-analyzer - **GitHub**: https://github.com/IleshDevX/TalentIQ - **Technologies**: Python, FastAPI, Streamlit, Sentence-Transformers, NLP, PyPDF2 - **Problem**: Legacy ATS parsers rely on rigid keyword matching, failing to capture semantic qualifications and failing to diagnose candidate skill gaps. - **Solution**: NLP semantic scoring engine using Sentence-Transformers to calculate dense vector cosine similarity between resumes and target job profiles with radar chart gap diagnostics. - **Outcome**: Sub-200ms semantic inference and multi-format PDF/DOCX structured text parsing. ### 2. LICS — Logistics Intelligence & Command System - **Category**: Logistics & Operational AI - **URL**: https://ilesh.dev/projects/lics-logistics-command-system - **GitHub**: https://github.com/IleshDevX/Logistics_Intelligence_Command_System - **Technologies**: Python, Streamlit, Scikit-learn, Decision Trees, Pandas, Weather API - **Problem**: In-transit delivery failures and cargo damage due to unforeseen weather conditions and unvetted route bottlenecks before dispatch. - **Solution**: Human-in-the-loop control tower combining Decision Tree ML classification with live WeatherAPI telemetry and multi-role governance (Seller, Manager, Supervisor). - **Outcome**: Enabled operational pre-dispatch risk assessment with supervisor override auditing. ### 3. Stock Trend Prediction & Financial Analytics - **Category**: Deep Learning & Quantitative Finance - **URL**: https://ilesh.dev/projects/stock-trend-lstm-prediction - **GitHub**: https://github.com/IleshDevX/Stock-Trend-Prediction - **Technologies**: Python, Streamlit, TensorFlow, Keras, LSTM, Plotly, yfinance - **Problem**: Time-series financial forecasting models suffer from look-ahead bias and fail to incorporate market sentiment into technical price trends. - **Solution**: Multi-layer LSTM neural network trained on scaled OHLC sequences paired with VADER NLP sentiment scoring from real-time financial news feeds. - **Outcome**: Interactive multi-timeframe price projection charts with 50-day and 200-day moving average indicators. ### 4. PocketQuant — Merchant Liquidity Shortage Prediction - **Category**: Quantitative Risk & MLOps - **URL**: https://ilesh.dev/projects/pocketquant-liquidity-prediction - **GitHub**: https://github.com/IleshDevX/PocketQuant - **Technologies**: Python, XGBoost, FastAPI, Streamlit, SHAP, Scikit-learn, Pandas - **Problem**: Sudden merchant liquidity defaults in high-volume ecosystems with extreme class imbalance (1.29% event rate). - **Solution**: Cost-sensitive XGBoost classifier with custom liquidity buffer ratios, SHAP feature attributions, and an operational risk threshold (0.40). - **Outcome**: Achieved 0.9956 accuracy, 0.844 F1 score, and 0.9991 ROC-AUC on 50,000 merchant-day records with a FastAPI prediction endpoint. ### 5. Pocket Planner — Personal Finance & Receipt Capture - **Category**: Full-Stack Intelligent Systems - **URL**: https://ilesh.dev/projects/poket-planner-ai-finance - **GitHub**: https://github.com/IleshDevX/POKET-PLANNER - **Technologies**: React, TypeScript, Node.js, Express, MongoDB, Tailwind CSS, Zod - **Outcome**: OCR receipt scanning into structured JSON transactions with JWT authentication and automated email reports. ### 6. Digital Mindset Prediction Tool - **Category**: Applied Machine Learning - **URL**: https://ilesh.dev/projects/digital-mindset-lasso-regression - **GitHub**: https://github.com/IleshDevX/Digital_Mindset - **Technologies**: Python, Streamlit, Scikit-learn, Lasso Regression, Plotly, Joblib - **Outcome**: L1-regularized Lasso Regression evaluating 30 organizational metrics for digital transformation readiness. ### 7. CodeOrbit Hub — Web App & API Collection - **Category**: Web Engineering & API Integrations - **URL**: https://ilesh.dev/projects/codeorbit-hub-apis - **GitHub**: https://github.com/IleshDevX/CodeOrbit_Hub - **Technologies**: JavaScript, HTML5, CSS3, REST APIs, Vanilla JS, EmailJS - **Outcome**: 35+ integrated web applications across 7 categories with client-side live search and external API data feeds. ### 8. EduRecruit — Campus Placement Portal - **Category**: Full-Stack Web Engineering - **URL**: https://ilesh.dev/projects/edurecruit-campus-portal - **GitHub**: https://github.com/IleshDevX/EduRecruit - **Technologies**: React, Node.js, Express, MongoDB, Tailwind CSS, Framer Motion, JWT - **Outcome**: Role-based portals for Students, Recruiters, and TPOs with automated eligibility criteria filtering. ### 9. Nexus Café — Futuristic Coffee Shop Showcase - **Category**: Interactive Front-End Design - **URL**: https://ilesh.dev/projects/nexus-cafe-experience - **GitHub**: https://github.com/IleshDevX/Nexus-Cafe - **Technologies**: HTML5, Tailwind CSS, JavaScript, GSAP, ScrollTrigger, FontAwesome ### 10. Insight Visions — Photography Gallery Web App - **Category**: Front-End Web Development - **URL**: https://ilesh.dev/projects/insight-visions-gallery - **GitHub**: https://github.com/IleshDevX/Insight_Visions - **Technologies**: HTML5, CSS3, JavaScript, Tailwind CSS, FontAwesome --- ## Professional Work Experience & Research ### AI/ML Associate Intern — Technetics IT Services (Jan 2026 — Present) - Developed RAG document question-answering systems using dense vector embeddings. - Implemented text preprocessing, semantic chunking, and vector search. - Built Variational Autoencoders (VAEs) for feature learning and anomaly detection. - Deployed trained machine learning models into production FastAPI REST endpoints. ### Research Intern (Hardware Team Lead) — Logicrest (Nov 2025 — Jan 2026) - Led IoT hardware development for an autonomous vehicle telemetry system. - Designed sensor interface circuits and low-latency serial data communication. - Implemented edge data acquisition firmware and collaborated on telemetry pipelines. --- ## Education & Academic Background - **Degree**: Bachelor of Engineering (B.E.) in Information Technology - **Institution**: C.K. Pithawalla College of Engineering & Technology (GTU) - **Period**: 2023 — 2027 - **Minor Degree**: Data Science - **Focus**: Data Structures & Algorithms, Artificial Intelligence, Machine Learning, Deep Learning, NLP, Database Management Systems, Computer Networks. --- ## Professional Services & Offerings 1. **Generative AI & RAG Pipeline Engineering**: Context-aware document Q&A engines, dense vector search, and hallucination guardrails. 2. **Production Machine Learning & Deep Learning**: Supervised/unsupervised models, time-series forecasting, and SHAP explainability. 3. **Full-Stack Intelligent Web Applications**: Next.js/React web applications powered by AI inference microservices. 4. **IoT Anomaly Detection & Edge ML**: Variational Autoencoders (VAEs) and sensor telemetry monitoring. 5. **Quantitative Risk & Financial Modeling**: Liquidity risk scoring, trend forecasting, and financial news sentiment analysis. 6. **Model Serving & Low-Latency API Services**: FastAPI/Docker microservices with sub-100ms response times. --- ## Canonical Portfolio Navigation - **Homepage**: https://ilesh.dev/ - **About**: https://ilesh.dev/about - **Projects**: https://ilesh.dev/projects - **Experience**: https://ilesh.dev/experience - **Blog & Technical Notes**: https://ilesh.dev/blog - **Resume**: https://ilesh.dev/resume - **Contact**: https://ilesh.dev/contact