Research & Material Science Project

AstraMix AI

AI-assisted sustainable concrete mix design and optimization platform.

Created by Nikolay Khachatryan

Predict Strength

Run machine learning evaluations to forecast 28-day concrete strength using raw ingredients.

Optimize Mix

Search for optimal low-carbon and cost-effective mixes using our bounded SLSQP solver.

Model Comparison

Explore the validation metrics (RMSE, MAE, R²) across XGBoost, Random Forest, and baseline models.

Report Preview

Read the project report summary covering problem motivation, features, and model validations.

About Research

Learn more about AstraMix's methodology, material limits, and Nikolay Khachatryan's work.

Backend Docs

Explore the interactive Swagger/OpenAPI schemas, schemas, and live test endpoints.