Safiullah Rattar

Software Engineer — AI Infrastructure & Full‑Stack

RAG pipelines, vector search, real-time systems, and production web apps.

Specimen Archive

The Workbench

Parhako

EdTech Platform

SPECIMEN №001

AI study platform with RAG pipelines, structured output parsing, mock tests, and exam workflows. 2,500+ active users. Processes 20,000+ practice questions. Built with Next.js, PostgreSQL, LangChain, AWS ECS, S3, and SES.

View Project →

vectordb-lite

Vector Database

SPECIMEN №002

HNSW vector database built from scratch in Python + NumPy with multi-layer graph construction, cosine/euclidean/dot-product similarity, and soft-delete persistence.

View on GitHub →

shaderlab

Creative Coding

SPECIMEN №003

Creative coding platform for GLSL shaders with browser-based live preview, real-time compilation, and a gallery workflow for sharing shader experiments.

View on GitHub →

Service Record

Contract Work

Trading data pipeline. Automated news gathering, market scanning, statistical analysis for real-time trading. Python, WebSocket, PostgreSQL, AWS ECS. Circuit breakers, risk checks, position tracking.

University LMS. Course delivery, assignments, automated grading, progress dashboards. React, Node.js, PostgreSQL, Docker.

Medical practice platform. Embedded patient management widget for healthcare SaaS. Records, appointments, clinical workflows. React, Node.js, PostgreSQL.

Warehouse terminal. Legacy terminal interfaces rebuilt in Angular, Java/Spring Boot. Real-time WebSocket sync, offline-capable mobile interfaces.

Contractor risk app. Background checks, QuickBooks financial data integration, lead scoring. TypeScript, QuickBooks API, PostgreSQL.

Carleton University. B.Eng, Software Engineering. GPA 9.98 / 12.00.

Technical Inventory

The Chalkboard

e^iπ + 1 = 0

∇ × B = μ₀J + μ₀ε₀ ∂E/∂t

F = ma

∫₀^∞ e^−x² dx = √π/2

Skills & Competencies

Python

TypeScript

React

Next.js

Node.js

PostgreSQL

AWS

Docker

CI/CD

Linux

LLMs

RAG

Embeddings

Vector Search