Abdullrahman
HomeAboutProjectsResearchSkillsContact
Download CV
Back to Archive
Project 01 // Core Engine

I Built a Mini Palantir Foundry-Style Platform in FastAPI and
Next.js.

An open-source, self-hosted data platform inspired by Palantir Foundry

VIEW ON GITHUB
FastAPI
FAS
FastAPI
120+Backend API Endpoints
30+Frontend Pages
12+Governance Capabilities
Process // 01

The Challenge

Modern data stacks are powerful, but often fragmented. Teams use separate tools for ingestion, transformations, dashboards, notebooks, permissions, AI, workflows, and operations. Each tool may work well alone, but the gaps between them create serious problems: permissions drift, lineage becomes incomplete, data gets copied, governance is inconsistent, and AI can accidentally bypass security controls. Mini Foundry was built to explore a compact Foundry-style alternative: one self-hosted platform where data can be connected, cataloged, governed, queried, transformed, modeled as business objects, used in dashboards and applications, assisted by AI, and audited through one shared backend permission and resource model. The challenge was not simply building individual features like a CSV uploader, SQL editor, dashboard builder, notebook, or job queue. The harder challenge was making sure every feature obeys the same governance rules at execution time.

Bottleneck Analysis

Identification of relay failures in distributed nodes.

Process // 02

The Solution

Advanced Core Logic

One place to connect data, govern it, transform it, model it as business objects, and build analytics or operational apps on top.

Backend: FastAPI with all feature routers mounted under /api/v1.
Frontend: Next.js App Router, React, TypeScript, platform shell, command palette, role-aware navigation, and branch selector.
bash
cp .env.example .env
docker compose up -d --build
http://localhost:3000
Email: admin@mini.local
Password: admin
http://localhost:8000/docs
http://localhost:8000/api/v1/system/health
Architecture

Visual Chronicles

Abdullrahman

© 2026 Abdullrahman. Built for high-end AI research.

LinkedInGitHubPrivacyTerms
Built a self-hosted FastAPI and Next.js platform that connects ingestion, catalog, SQL, pipelines, ontology, dashboards, notebooks, AI, jobs, audit logs, and operations into one unified product.
  • Created a shared platform kernel around resources, capabilities, ResourceACLs, markings, versions, branches, approvals, exports, and audit events so datasets, pipelines, dashboards, apps, and models can be governed consistently.
  • Implemented governed SQL using sqlglot validation, dataset reference resolution, capability checks, row policies, column masks, query rewriting, policy-aware caching, and audit logging.
  • Solution visualization