Hey there!
I'm Yash β a software engineer who builds things that matter. From backend systems handling real-world data to on-device ML that runs in your pocket, I love turning complex problems into elegant solutions. If you're looking to collaborate or just want to chat about tech, let's connect π
<Yash Salunke />
Software Engineer Γ ML Engineer
Building scalable backend systems, intelligent ML solutions, and native iOS apps with precision engineering.
About Me
Hello, Iβm Yash Salunke, a software engineer who enjoys building dependable, user-facing systems that sit at the intersection of backend engineering, mobile development, applied machine learning, and data science. Iβm motivated by problems that require thoughtful system design, strong data foundations, and practical AI integration.
My experience spans both production systems and research-driven projects, where Iβve focused on building scalable services, reliable pipelines, and intuitive user experiences. I care deeply about clean architecture, maintainability, and designing systems that are easy to reason about as they grow in complexity.
Much of my work has involved translating ideas into working software from backend services and APIs to mobile interfaces and data-driven components. I enjoy working across layers of the stack and collaborating closely with others to turn ambiguous requirements into robust, well-tested systems.
Outside of core development, Iβm particularly interested in how data and machine learning interact with software systems in real-world settings β whether thatβs improving performance, observability, or decision-making. I value continuous learning and enjoy building things that have tangible impact beyond code.
β what i do best
Core Expertise π―
Backend & Systems Engineering
I design and build scalable backend systems with a focus on correctness, clean APIs, and real-time data handling. My work spans distributed systems, microservices, and cloud infrastructure.
Machine Learning & Applied AI
I apply machine learning to solve practical problems across healthcare, computer vision, and NLP. I care about model evaluation, interpretability, and data quality.
Mobile & On-Device Systems
I build native iOS applications with emphasis on performance and real-time processing. My work integrates sensors, on-device ML workflows, and backend systems.
Cloud Architecture & DevOps
I architect cloud solutions on AWS and Azure with automated CI/CD pipelines. From containerization to infrastructure-as-code, I ensure reliable deployments.
Research & Data Engineering
I build systems where engineering decisions impact research outcomes. From designing experiments to optimizing pipelines, I focus on technical choices that support analysis.
Full-Stack Development
I develop end-to-end solutions with modern tech stacks including React, Node.js, and Python. My work bridges frontend experiences with robust backend architectures.
β check out some
Projects & Experience π»
Hazard Perception Research Platform
Johns Hopkins University
Tech Stack: Azure, Java, Python, CI/CD
- Cloud research platform for behavioral studies
- Designed backend experiment logic and data pipelines
- Implemented real-time analytics and automated reporting
360Β° Car Inspection System
CCC Intelligent Solutions
Tech Stack: SwiftUI, Core Motion, AWS Bedrock, Docker
- iOS pipeline with on-device GenAI for real-time inspection
- Implemented ML-driven frame quality assessment
- Integrated with cloud-based processing system
π Internal / Private Project
Metabolic Syndrome Prediction
Research Publication
Tech Stack: Python, XGBoost, SHAP, Machine Learning
- ML study across 9 models on 2,400 patients
- XGBoost achieved 95% training and 88.97% testing accuracy
- Published in Procedia Computer Science (Elsevier)
LLM-BPE-BERT
Academic Project
Tech Stack: Python, PyTorch, NLP, Transformers
- Custom BERT model with BPE tokenizer
- Implemented from scratch for NLP tasks
- Achieved competitive performance on benchmark datasets
VoiceYourID
Security Project
Tech Stack: Python, GMM, MFCC, React, PWA
- Voice recognition system with 88% accuracy
- Implemented Gaussian Mixture Models for speaker identification
- Built Progressive Web App for cross-platform deployment
Real-Time Options Chain
Market Analysis Engine
Tech Stack: Java, Node.js, Angular, WebSockets, REST APIs
- Real-time market data processing and analysis
- WebSocket-based live streaming architecture
- Full-stack financial analytics platform
β research
Publications π
Metabolic Syndrome Prediction
Procedia Computer Science (Elsevier) ICMLDE 2023, Volume 235, 2024Evaluated 9 machine learning classifiers on 2,400-patient dataset. XGBoost achieved 95% training and 88.97% testing accuracy with 0.913 F1 score, identifying waist circumference as the most predictive biomarker.
Smart City IoT Research
IEEE Conference Publication 2024Research on smart city technologies and Internet of Things applications. Focused on developing intelligent systems for urban infrastructure and connectivity.
β career
Work Experience πΌ
Research / Software Engineer
Johns Hopkins University Jan 2024 β Present- Built and maintained a cloud-based research platform for driving hazard perception studies
- Designed backend experiment logic and data pipelines for high-fidelity behavioral data
- Implemented real-time data processing and analysis workflows using Azure and Python
Data Science Intern
CCC Intelligent Solutions May 2023 β Dec 2023- Built a real-time iOS inspection pipeline using on-device GenAI (FastVLM)
- Implemented frame extraction, blur filtering, and ML-driven retake workflows
- Developed SwiftUI applications with Core Motion integration and AWS Bedrock
Data Analyst Intern
Nasdaq Jan 2024 β Jun 2024- Developed automated data pipelines and forecasting models for market analysis
- Built dashboards that significantly reduced reporting time
- Won Nasdaq Hackathon for developing innovative trading solutions
Data Science & Analytics Intern
Six Ladders Pvt. Ltd. Jun 2023 β Aug 2023- Built analytics dashboards and churn prediction models
- Improved data retrieval performance and user retention
- Implemented Python-based ETL pipelines for data processing
β let's connect
Get In Touch π¬
Open to new opportunities and collaborations. Feel free to reach out!