I'm Saptarshi Sadhu
Full Stack Developer & ML Researcher
|
I focus on designing and engineering scalable full-stack systems, intelligent applications, and robust software architectures where performance, structure, and real-world usability matter.
About Me
Designing systems where software, data, and intelligence converge.
The Foundation
I am a Computer Science undergraduate focused on understanding and building complex digital systems. My academic journey has been driven by curiosity about how software, data, and infrastructure interact to create scalable and intelligent applications. This focus is reflected in my academic performance, maintaining CGPAs of 9.3, 9.5, and 9.43 across semesters.
Algorithms
I maintain a strong focus on algorithmic thinking and problem solving, earning 4★ in C, 3★ in C++, and 3★ in Problem Solving on HackerRank.
System Architecture
My work spans full-stack development, algorithmic problem solving, and machine learning research. I enjoy designing systems end-to-end — from frontend interfaces and backend services to databases and infrastructure — ensuring that each layer works cohesively within a larger architecture.
Tech Ecosystems
I actively strengthen my technical foundation through continuous learning and experimentation across modern technologies including JavaScript ecosystems, Python, React, Node.js, databases, cloud platforms, and AI frameworks. Rather than treating tools as isolated skills, I focus on understanding how technologies integrate to form complete software ecosystems. This includes how frontend frameworks interact with backend services, how data flows through applications, and how infrastructure supports scalable deployments. Through structured learning, project development, and experimentation, I continuously expand my exposure to different programming paradigms, frameworks, and architectural patterns. This approach helps me build a deeper understanding of the broader software landscape rather than limiting myself to a single stack.
Machine Learning Research
Alongside software development, I am actively involved in ML research on urban air quality prediction, exploring models that analyze pollutants (VOCs, O₃, NOx) using environmental and temporal data. The project investigates how data-driven techniques can help understand complex patterns and support better forecasting.
Long-Term Vision
My long-term goal is to grow into an engineer capable of designing intelligent systems, building scalable software infrastructure, and contributing to technologies that operate at real-world scale.
Key Highlights
Education
WBBSE
WBCHSE
CGPA (Current)
Coding
HackerRank
Research
(VOCs, O₃, NOx)
Engineering Experiments
Authentication Flow Prototype
Interactive Portfolio Interface
Frontend Component Experiments
Backend API Practice
Focus
Technical Proficiency
Skills & Tech Stack
terminal Programming Languages
core · systems · mobile
language Web Technologies
frontend · frameworks · ui
database Databases & BaaS
storage · realtime · persistence
dns Backend Frameworks
apis · servers · microservices
handyman Tools & Platforms
dev environment · productivity
cloud Cloud Infrastructure
deploy · scale · distribute
science Machine Learning Research
Urban Air Quality Prediction · VOCs, O₃, NOx
track_changes Current Focus
AI × backend engineering × scalable systems
Selected Projects
C & Systems
low-level & performance
Custom Memory Allocator
A high-performance thread-safe malloc implementation with minimal fragmentation using segregated fits.
TCP/IP Stack Simulation
Network packet processing engine for handling custom transport layer protocols in user space.
Flutter & Mobile
cross-platform UI
Focus Flow
Minimalist productivity tracker featuring smooth 60fps animations and offline-first architecture.
Wealth Watch
Cryptocurrency portfolio manager with real-time websocket updates and custom chart rendering.
Research & Publications
Optimizing Distributed Inference in Latency-Critical Edge Environments
Investigated novel partitioning strategies for deep neural networks across heterogeneous edge nodes. The proposed model reduces communication overhead by 34% while maintaining model accuracy in dynamic network conditions.
Predictive Cache Management for High-Frequency I/O Operations
Developed a heuristic-based prefetching algorithm for non-volatile memory architectures. This research demonstrates a significant reduction in tail latency for database workloads under heavy concurrent access.
Academic Background
Education
Consistent academic performance with strong technical foundation.
Secondary Education
B. T. Road Gov. Spon. H. S. School
Higher Secondary
B. T. Road Gov. Spon. H. S. School
B.Tech in Computer Science
Adamas University
Open to Opportunities
I’m currently looking for opportunities where I can contribute, learn, and build real-world systems. I enjoy working on backend development, problem solving, and scalable applications.
Track Record
Achievements & Engagement
Highlights from hackathons, continuous learning, and technical milestones.
Technical Achievements
- Built a complete full-stack auth system solo, start to finish
- Trained an ML model on real-world air quality data (VOC, O₃, NOx)
- Shipped a Unity horror game prototype with custom AI enemy logic
- Consistently moved from idea to working prototype without extended planning
Hackathons & Competitions
- Urban Vision Hackathon 2025 — Ministry of Education, India
- Smart India Hackathon (SIH) — national-level participation
- Adobe India Hackathon — design & tech challenge
- MH Techin Innovation Challenge — cleared multiple rounds
- Hackground India 2K25 & Ideat-A-Thon via Unstop
Coding Engagement
- Solved structured problems in #T20DSA Challenge — Stacks, Queues & Range Queries
- Actively competed in CodeClash August Arena timed challenge
- Competed in Clash of Coders 2.0 and 3.0 — finalist in 2.0
- Engaged in DSA Quiz & Coding Competitions across Unstop
Events & Workshops
- Competed in 24-Hour UI/UX Design Hackathon using Figma
- Attended React.js Unfiltered — ALCHEMIST workshop
- Engaged in IdeateX ideation & product thinking sessions
- Participated in BuildFolio — portfolio development event
Aptitude & Assessments
- Appeared in technical & aptitude quizzes on Unstop
- Cleared multiple rounds of MH Techin Innovation Challenge
- Participated in cross-domain tech challenges beyond core stack
- Consistently active across competitive platforms throughout college
Recognition & Exposure
- Shortlisted for AWS College Showcase Program
- Selected for CTRLS Data Center Industrial Visit
- Participated in Smart India Hackathon — national scale
- Represented college at Clash of Coders competition
Let's Connect
Open to collaborations, projects, and meaningful conversations.
Let’s build something meaningful.