Computer Science Engineer - Full Stack & AI Systems Developer

Tejas Mahakalkar

House no.***, Mathipura, Hanuman Nagar, Nagpur

755-***-**** # ******************@*****.*** ï Tejas Mahakalkar § Tejas3104 Profile

Final-year Computer Science engineer with hands-on experience building full stack applications and AI-powered systems. Skilled in Python, Node.js, React, and REST API development. Actively use AI tools (Claude, GitHub Copilot, ChatGPT) daily to accelerate development, debug issues, and ship features faster. IEEE-published researcher. High-ownership mindset with experience taking projects from idea to deployment. Open to relocation to Mumbai. Technical Skills

Frontend: React, Next.js, HTML, CSS, JavaScript

Backend: Python, Node.js, REST APIs, Server-side Development, Flask Database: PostgreSQL, MongoDB, MySQL, Firebase Realtime Database AI/ML: PyTorch, TensorFlow, Keras, Scikit-learn, YOLOv8, OpenCV, NumPy, Pandas Tools & DevOps: Git, GitHub, Linux, VS Code, Google Cloud Platform, CI/CD (basics) AI Dev Tools: Claude, GitHub Copilot, ChatGPT — used daily for code generation, debugging, and development acceleration

Education

Shri Ramdeobaba College Of Engineering and Management Jan 2023 – Present Bachelor of Engineering in Computer Science — CGPA: 8.15 Nagpur, Maharashtra Government Polytechnic, Nagpur Dec 2020 – Jul 2023 Diploma in Computer Engineering — Percentage: 92.70% Nagpur, Maharashtra Internship

iBase Electrosoft LLP Jun 2022 – Jul 2022

Software Development Intern Nagpur, Maharashtra

• Built a Python-based audiobook application with modular backend architecture, improving audio processing efficiency by 25%.

• Performed end-to-end debugging and deployment validation across 20+ test cases, reducing pre-release defects by 15%.

Projects

Image Based Forest Analysis & Optimal Path Computation Python, YOLOv8, OpenCV, ML GitHub 2024 – 2025

• IEEE-published AI-powered system analyzing satellite imagery to detect tree crowns, estimate green cover, and compute optimal traversal paths.

• Fine-tuned a YOLOv8 model achieving 77.1% precision and 76.6% recall using transfer learning on a custom annotated dataset.

• Engineered a modular backend with three integrated pipelines — tree detection, vegetation segmentation, path planning — exposed via REST APIs.

EcoSort Python, TensorFlow, Keras, Streamlit, OpenCV, REST APIs GitHub Sep 2024

• AI-powered waste classification using VGG16 and ResNet50 deep learning models, improving accuracy from 71% to 90% via transfer learning and fine-tuning.

• Built fullstack inference pipeline: backend preprocessing (500+ images) integrated into a Streamlit UI, delivering real-time predictions under 2-second latency.

• Designed backend as a deployable ML service with modular preprocessing, model inference, and output rendering — end-to-end feature ownership from model to UI.

HarvestPro HTML, CSS, JavaScript, Python, ML (Random Forest), REST APIs GitHub Feb 2025

• Built a fullstack smart agriculture platform integrating crop/fertilizer recommendation, disease detection, and AgriBot

(AI chatbot) — serving 500+ dataset entries.

• Achieved 99.3% crop recommendation accuracy and 96.4% fertilizer accuracy using Random Forest models trained on curated datasets.

• Designed backend REST APIs consumed by the frontend; integrated AgriBot via OpenRouter.ai external API

— demonstrating third-party API integration, authentication, and async workflows.

Contact this candidate