Perumalla Satya Varun Profile Picture

Hi, I’m Perumalla Satya Varun

Aspiring Software Developer | Web β€’ Machine Learning β€’ Python β€’ Full Stack

Summary & Overview

Aspiring software developer with a strong foundation in web technologies and programming. Proficient in HTML, CSS, JavaScript, Python, Java, SQL, and PHP with hands-on experience in front-end and back-end development, database management, and deep learning model training.

Eager to contribute to the development of scalable, user-centric applications in a dynamic and growth-oriented environment that values efficiency, creativity, and technical excellence.

Engineering Principles

How I approach software design, problem solving, and system architecture

⚑

Clean & Scalable Code

Writing maintainable, modular code adhering to modern design patterns and best practices.

🧠

Algorithmic Efficiency

Analyzing time & space complexity to build high-performance data processing pipelines.

🎨

User-Centric Design

Designing modern, responsive interfaces with smooth micro-animations and intuitive navigation.

πŸš€

Continuous Innovation

Exploring emerging AI models, audio signal processing, and modern full-stack frameworks.

Education

Vignan’s Foundation for Science Technology and Research

2022 – 2026

πŸŽ“ Bachelor of Technology in Information Technology

CGPA: 8.15 | Guntur, India

Jaya Bharathi Junior College

2020 – 2022

🏫 MPC (Class XII)

Percentage: 90.03% | Gurazala, India

Technical Skills

πŸ’» Programming Languages

PythonC++CJava

🌐 Web Technologies

JavaScriptCSSHTMLReact.jsPHPTailwind CSS

🧠 AI & Machine Learning

Deep Learning (CNN, VGG19, Inception)PyTorchTensorFlowIntroductory exposure to LLMs

βš™οΈ Backend & APIs

DjangoFastAPIFlaskREST APIsJSONNode.jsExpress

πŸ“Š Python Libraries

NumPyPandasMatplotlibScikit-learnLibrosa

πŸ›’οΈ Databases

SQLMongoDB

☁️ Cloud Platforms & Tools

AWS (EC2, S3)GitHubVS CodeGoogle ColabXAMPPGit

Featured Projects

Full-Stack MERN & AI Fitness Platform

Fit Pulse

Comprehensive full-stack fitness application using the MERN stack with secure JWT authentication and role-based dashboards for Users, Trainers, and Admins.

  • ⚑ Integrated TensorFlow.js and Google Generative AI for real-time exercise posture tracking and predictive fitness analytics.
  • ⚑ Built real-time networking and instant messaging features using Socket.io between members and trainers.
  • ⚑ Designed responsive UI using Tailwind CSS, Framer Motion, and Recharts for interactive workout data visualization.
React.jsNode.jsExpressMongoDBJWTTensorFlow.jsSocket.ioTailwind CSSFramer Motion
Python & Deep Audio Processing

Query-by-Humming Music Retrieval

Humming-based music identification system using Python, Librosa, and deep audio processing techniques to retrieve songs from hummed queries.

  • ⚑ Implemented pitch extraction and normalization using PYIN algorithm with DTW for accurate song matching.
  • ⚑ Utilized Audio Spectrogram Transformer (AST) and LAION-CLAP models for robust audio embeddings.
  • ⚑ Tested on a Telugu melody dataset achieving effective Top-K song retrieval with scalable support for large databases.
PythonLibrosaPYINDynamic Time Warping (DTW)AST TransformerLAION-CLAPMel Spectrograms
Computer Vision & Deep Learning

Goat Disease Identification Using Deep Learning

Deep learning disease classification system trained on a custom dataset to identify livestock health issues accurately.

  • ⚑ Applied image augmentation techniques, increasing training data volume by 60%.
  • ⚑ Trained CNN, VGG19, and Inception models, achieving 85% classification accuracy.
  • ⚑ Improved overall model diagnostic performance by 10–15% post augmentation.
PythonCNNVGG19InceptionPyTorchTensorFlowOpenCV

Code Showcase & Engineering Logic

Interactive look at my algorithm implementations & clean architecture

Audio Pitch Matching (PYIN & DTW)
import librosa
import numpy as np

def extract_pitch_contour(audio_path, sr=22050):
    y, sr = librosa.load(audio_path, sr=sr)
    # PYIN algorithm for fundamental frequency (f0)
    f0, voiced_flag, voiced_probs = librosa.pyin(
        y, fmin=librosa.note_to_hz('C2'), fmax=librosa.note_to_hz('C7')
    )
    # Normalize pitch vector for Dynamic Time Warping matching
    f0_clean = np.nan_to_num(f0)
    return f0_clean / (np.max(f0_clean) + 1e-6)

πŸ’‘ Core pitch extraction module used in the Query-by-Humming music identification system.

Achievements

  • ⭐ 2* Rating in CodeChef (2024)
  • ⚑ 200+ Coding Problems Solved: Solved 200+ problems across various coding platforms including LeetCode, CodeChef, and HackerRank.
  • πŸ† National-Level Hackathons: Competed in multiple national-level hackathons including HackwithInfy and Amazon Hackathon, gaining hands-on problem-solving experience.

Courses & Certifications

πŸ“œ NPTEL Certified in Programming in C++ (2024)
πŸ“œ HackerRank Python (Basic) Certification
πŸ“œ HackerRank SQL (Basic) Certification

Frequently Asked Questions

Quick insights into my background, role availability, and engineering approach

Are you available for full-time software developer roles or internships?

βˆ’

Yes! I am actively looking for full-time software development roles and internships in Full-Stack Web Development, Python Engineering, and AI/Machine Learning upon my 2026 graduation.

What are your core technical strengths?

+

What key projects have you built independently?

+

How do you handle team collaboration and version control?

+

Get In Touch

Feel free to reach out for software engineering opportunities, collaborations, or technical projects!

πŸ“± Phone

+91 8247485071

βœ‰οΈ Email

perumallavarun171@gmail.com