I'm a passionate full-stack developer and software engineer with hands-on experience building production-grade, scalable, and secure applications across web and mobile platforms. My expertise spans full-stack web development, Android app development (Kotlin/Jetpack Compose), machine learning pipelines, and cloud deployments — with a strong emphasis on clean architecture, secure backends, and seamless integrations using modern frameworks and cloud services.
I've authored two research publications, built AI-driven HR analytics systems with Explainable AI (XAI/SHAP), engineered a serverless decentralized Android messaging app using BLE Mesh and Tor, and developed real-world enterprise systems under Agile/JIRA methodology with CI/CD pipelines.
A curious and hands-on software engineer with experience in full-stack web development, Android development, machine learning, and cloud-native systems. Passionate about crafting secure backends, intelligent ML pipelines, interactive UIs, and scalable architectures using modern frameworks and cloud platforms.
Ailexity Software – Software Development InternPune, IN (Hybrid/Remote) | June – July 2025
Technologies: Spring Boot, React.js, JWT, Hibernate ORM, MySQL, React Query, Context API, TestNG, JIRA, Agile.
MIT ADT University, PuneB.Tech – Information Technology | 2023 – 2027
CGPA: 7.35
HSC (12th Grade) – Maharashtra State Board | 2021 – 2023
Percentage: 66.17%
SSC (10th Grade) – Maharashtra State Board | 2020 – 2021
Percentage: 90.40%
College Resource Sharing Platform
PWA for students to share resources, join study rooms, and manage notes. Firebase Auth (email + Google OAuth), RBAC, Firestore full-text search, real-time chat & voice rooms, push notifications, and offline-first Service Worker support.
Stack: React.js · Firebase Auth · Firestore · Firebase Realtime DB · EmailJS
View ProjectDecentralized Secure Chat Application
Serverless Android app — BLE Mesh for offline multi-hop relay, Tor (Arti/JNI) for anonymized onion-routed traffic, X25519 ECDH + AES-GCM E2E encryption via Google Tink, MVVM + Dagger-Hilt, Room DB. Sub-50ms response times with zero server dependency.
Stack: Kotlin · Jetpack Compose · MVVM · BLE Mesh · Tor (Arti/JNI) · Room DB
View ProjectEmployee Attrition Prediction System
ML pipeline benchmarking 4 classifiers (LR, SVM, RF, XGBoost) on IBM HR Analytics dataset; SMOTE for class-imbalance — AUC-ROC 0.8170, F1: 0.438, Accuracy: 65.2%. SHAP Explainability (XAI) for per-prediction attribution. Flask REST API with batch processing; deployed on AWS EC2 + S3 with Docker & GitHub Actions CI/CD.
Stack: Python · Flask · scikit-learn · XGBoost · SHAP · SMOTE · Plotly · Docker · AWS
View Project
Compact demo of the ShopLane e‑commerce app — product listing, cart and checkout preview.
ViewVenue: International Journal of Computer Applications (IJCA)
Author: Anurag Bodkhe (Corresponding Author)
DOI: https://doi.org/10.5120/ijca1f36c9153a77
ORCID: 0009-0007-2148-7777
Developed an XGBoost-based ML model with comparative analysis of Logistic Regression, SVM, Random Forest, and XGBoost on the IBM HR Analytics dataset (1,470 records, 35 features). Applied SMOTE for class-imbalance correction — achieving AUC-ROC 0.8170 and 65.2% accuracy. Integrated SHAP per-prediction feature attribution to surface top attrition drivers (overtime, job satisfaction, tenure), enabling actionable HR workforce retention insights.
Venue: ScienceOpen — Hosted document / Research Paper
Author: Anurag Bodkhe (Corresponding Author)
DOI: https://doi.org/10.14293/PR2199.002312.v1
ORCID: 0009-0007-2148-7777
https://www.scienceopen.com/hosted-document?doi=10.14293/PR2199.002312.v1
Designed a decentralized Bluetooth MANET system for internet-free P2P messaging; dynamic multi-hop BLE routing with adaptive TTL for resilient delivery in offline and emergency low-connectivity environments. Hybrid end-to-end encryption (X25519 ECDH + AES-GCM) via Google Tink; Tor (Arti/JNI) for anonymized onion-routed traffic — validated against eavesdropping, MITM, and replay attacks in infrastructure-less scenarios.