S.
Portrait of Siddharth S

Siddharth S

Computer Science undergraduate specialising in Cyber Security, building full-stack, backend and AI-powered applications.

About

I have software engineering internship experience building scalable full-stack, backend, AI-powered and machine learning applications. My foundation is in Data Structures & Algorithms, DBMS, Operating Systems, Computer Networks and application security.

Experience

Software Development Intern

Jun 2026 – Aug 2026

Kalika Technologies LLP

  • Engineered features for a Continuous Glucose Monitoring (CGM) mobile app with real-time monitoring and historical data visualization.
  • Optimized the Bluetooth Low Energy (BLE) communication stack, improving packet parsing, reconnection handling and synchronization reliability.
  • Refined sensor calibration algorithms and the glucose-reading pipeline, converting raw ADC values to accurate readings with drift correction.
  • Validated algorithms against sensor logs and CSV datasets, debugged data-quality issues and added unit tests to improve reliability.

Projects

Tymely — AI-based Academic Timetable Generator

GitHub Live

React · JavaScript · Vite · Firebase Firestore · SheetJS · jsPDF · ESLint

  • Architected two metaheuristic scheduling engines (Genetic Algorithm and Ant Colony Optimization) across approx. 11,900 lines of JSX; ACO evaluated 3,200 candidate solutions per run.
  • Modeled 10+ hard and soft scheduling constraints aligned to NEP 2020 guidelines over a 112-slot timetable supporting multiple semester structures.
  • Built timetable simulation for teacher absence, enrollment growth and elective changes.
  • Added Excel/PDF export, Firestore synchronization, role-based authentication and Vercel deployment.

CapitalSense — Cash Flow Decision Intelligence System

GitHub Demo

Flutter · FastAPI · PostgreSQL · SQLAlchemy (async) · Alembic · Python · Tesseract OCR · Google Cloud Vision API · JWT · bcrypt · n8n

  • Developed a semi-autonomous fintech application that ingests invoices, receipts and financial documents to generate explainable cash-flow decisions.
  • Designed a deterministic payment-prioritization engine using penalty risk, due-date urgency, vendor flexibility and confidence scoring.
  • Built a two-tier OCR pipeline (Tesseract for printed documents, Google Cloud Vision API for handwritten receipts) feeding an async SQLAlchemy/Alembic data layer.
  • Implemented JWT access/refresh authentication with token rotation, bcrypt password hashing and SHA-256 hashing of sensitive fields (phone, GST, invoice IDs).
  • Automated vendor negotiation email drafting using n8n agentic AI workflows.

Converge — AI-based Campus Collaboration Platform

GitHub Live

React · TypeScript · Vite · Django · Spring Boot · PostgreSQL · TensorFlow · Vertex AI (Gemini) · Google Vision API · BigQuery

  • Engineered an AI-powered platform that matches students to teammates based on skills, interests and project requirements.
  • Built the React + TypeScript frontend (Vite) with resume upload, AI-powered skill extraction and teammate recommendations using Google Vision API and Vertex AI.
  • Collaborated with backend teams using Django, Spring Boot, PostgreSQL, and deployed the frontend on Vercel.

Poetry Insights — Containerised Full-Stack Poetry Platform

Live

Node.js · Express.js · MongoDB Atlas · JWT · Docker · Podman · Traefik

  • Developed a secure full-stack poetry platform with JWT authentication, RBAC and server-rendered UI.
  • Developed REST APIs with bcrypt authentication, ownership validation, MongoDB Atlas and schema validation.
  • Containerized the application using Docker/Podman and deployed behind Traefik with automatic HTTPS.

Telco Customer Churn Prediction

GitHub

Python · pandas · NumPy · scikit-learn · MLflow · matplotlib

  • Built an end-to-end binary classification pipeline on 7,032 customer records with 22 engineered features, using scikit-learn Pipelines and StandardScaler preprocessing.
  • Evaluated Logistic Regression, Random Forest and HistGradientBoosting with class-weight balancing for class imbalance, achieving ROC-AUC 0.8383 and 80.8% recall.
  • Performed 5-fold stratified cross-validation, threshold optimization and experiment tracking using MLflow.

LifeLedger — Consequence-Driven Financial Literacy Simulator

GitHub

Next.js · React · TypeScript · Tailwind · Zustand · Zod · Recharts · Gemini API · PWA

  • Built a decision-simulation PWA that teaches present bias by scheduling the consequences of a choice into future months rather than punishing it immediately.
  • Implemented a shadow-agent benchmark that replays a textbook policy through a byte-identical world (same seed, market path and constraints) plus a sub-millisecond counterfactual replay isolating one past choice out of 312 possible paths.
  • Designed an LLM layer that never computes: all figures are derived in TypeScript and passed as facts, while a response guard rejects any generated rupee amount absent from the request payload.

Skills

Languages
Python, Java, C++, C, JavaScript
Frameworks
React, Node.js, Express.js, FastAPI
Databases
PostgreSQL, MySQL, MongoDB, Firebase Firestore
Developer Tools
Git, GitHub, Docker, Podman, Linux, Postman, Traefik, Vercel
Security
JWT Authentication, bcrypt, SHA-256 Hashing, RBAC, HTTPS/TLS
Core CS
Data Structures, Algorithms, OOP, DBMS, Operating Systems, Computer Networks, REST APIs, Software Engineering
AI / ML
MLflow, Tesseract OCR, n8n, scikit-learn, pandas, NumPy, matplotlib

Education

Achievements & Leadership

Certifications