__  __           _      _ __   __         _ _
|  \/  | ___   __| | ___| |\ \ / /_ _ _ __| | |_
| |\/| |/ _ \ / _` |/ _ \ | \ V / _` | '__| | __|
| |  | | (_) | (_| |  __/ |  | | (_| | |  | | |_
|_|  |_|\___/ \__,_|\___|_|  |_|\__,_|_|  |_|\__|

AI/ML Model Registry and Deployment Platform

Version, deploy, and monitor your machine learning models at startup scale.

ModelVault is a production-ready ML model management platform for teams that need full control over their AI assets: register models, track datasets, orchestrate training jobs, and serve predictions via a secure inference API -- all from a single dashboard.

---

Why ModelVault?

The ML ecosystem is fragmented. Models live in S3 buckets, experiment results scatter across notebooks, and deploying a new version means SSH-ing into a server. ModelVault centralizes the entire ML lifecycle into one auditable, team-friendly platform.

---

Feature Highlights

Model Registry

• Register models from any framework: PyTorch, TensorFlow, scikit-learn, ONNX, HuggingFace, XGBoost
• Full semantic versioning with description and metadata per version
• File upload for model artifacts with SHA-256 checksum verification
• Public/private model visibility

Dataset Management

• Upload and catalog training datasets with automatic row/column statistics
• Link datasets to training jobs for full lineage tracking
• Storage-agnostic: local filesystem in dev, S3-compatible in production

Training Job Orchestration

• Queue and monitor CPU/GPU training jobs with real-time log streaming
• Cancel in-flight jobs with graceful cleanup
• Job status lifecycle: QUEUED -> RUNNING -> COMPLETED / FAILED / CANCELLED

Inference API

• Deploy any registered model version as a live prediction endpoint
• REST API: POST /api/v1/inference/{model_id}/predict
• API key authentication for CI/CD and external integrations
• Request logging with latency tracking

Team Management

• Multi-tenant with FREE / PRO / ADMIN roles
• Full API key management: create, list, revoke
• Per-user model and dataset ownership with visibility controls

---

Architecture

+--------------------------------------------------------------+
|                      CLIENT (Browser)                        |
|  React 18 - TypeScript - Vite - Tailwind CSS                |
|  Zustand (auth) - TanStack Query (server) - Recharts        |
+------------------------+-------------------------------------+
                         |
                         |  REST  +  Bearer JWT  /  X-API-Key
                         |
+------------------------v-------------------------------------+
|                    BACKEND (Python 3.11)                     |
|  FastAPI - SQLAlchemy 2.0 async - Pydantic - Alembic        |
|                                                              |
|  +--------+  +---------+  +----------+  +-----------+      |
|  |  Auth  |  | Models  |  | Training |  | Inference |      |
|  | Routes |  | Routes  |  |  Routes  |  |  Routes   |      |
|  +--------+  +---------+  +----------+  +-----------+      |
|                                         Celery Workers      |
+------------------------+-------------------------------------+
                         |
                    Async ORM (SQLAlchemy)
                         |
       +-----------------+------------------+
       |                                    |
+------v------+                     +-------v------+
| PostgreSQL  |                     |   Redis 7    |
|     16      |                     | (queue/cache)|
+-------------+                     +--------------+

---

Tech Stack

Layer	Technology	Purpose
Runtime	Python 3.11	Backend server
Framework	FastAPI	Async REST API
ORM	SQLAlchemy 2.0 (async)	Database layer
Database	PostgreSQL 16	Primary data store
Cache/Queue	Redis 7, Celery	Background jobs
Auth	JWT (python-jose) + bcrypt	Stateless auth
Frontend	React 18, TypeScript, Vite	UI application
Styling	Tailwind CSS	Dark-theme design
State	Zustand + TanStack Query	Client state + caching
Charts	Recharts	Dashboard metrics
Containers	Docker, Docker Compose	Service orchestration
CI	GitHub Actions	Build and lint on push

---

Quick Start

Option A: Docker (recommended)

# 1. Clone
git clone https://github.com/avase33/modelvault.git
cd modelvault

# 2. Configure
cp .env.example .env
# Edit .env -- set SECRET_KEY, DATABASE_URL, REDIS_URL

# 3. Start all services
docker compose up -d

Service	URL
Frontend	http://localhost:3000
API	http://localhost:8000
Swagger Docs	http://localhost:8000/docs

Option B: Local Development

Backend

cd backend
python -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install -r requirements.txt
cp ../.env.example .env
uvicorn app.main:app --reload

Frontend

cd frontend
npm install
npm run dev

---

API Reference

All endpoints under /api/v1. Auth: Authorization: Bearer <token> or X-API-Key: <key>.

Module	Method	Endpoint	Description
Auth	POST	/auth/register	Create account
	POST	/auth/login	Get JWT tokens
	GET	/auth/me	Current user profile
Models	GET	/models	Browse models
	POST	/models	Create model
	POST	/models/{id}/versions	Add version
	POST	/models/{id}/versions/{vid}/upload	Upload artifact
Datasets	GET	/datasets	List datasets
	POST	/datasets	Create dataset
	DELETE	/datasets/{id}	Remove dataset
Training	GET	/training	List jobs
	POST	/training	Create job
	GET	/training/{id}/logs	Stream logs
	POST	/training/{id}/cancel	Cancel job
Inference	POST	/inference/{model_id}/predict	Run prediction
API Keys	GET	/api-keys	List keys
	POST	/api-keys	Create key
	DELETE	/api-keys/{id}	Revoke key

---

Project Structure

modelvault/
+-- backend/
|   \-- app/
|       +-- core/            # Config, database, security
|       +-- models/          # SQLAlchemy ORM (User, MLModel, Dataset, TrainingJob)
|       +-- schemas/         # Pydantic request/response schemas
|       +-- api/v1/          # auth | models | datasets | training | inference | api-keys
|       \-- main.py          # App entry, CORS, routers
+-- frontend/
|   \-- src/
|       +-- pages/           # Dashboard, Models, Datasets, Training, API Keys
|       +-- components/      # Layout, shared UI
|       +-- store/           # Zustand auth store
|       \-- lib/             # Axios API client
+-- docker-compose.yml
+-- .env.example
\-- LICENSE

---

Roadmap

• Model playground -- browser-based live inference
• Experiment tracking with metric charts
• Python SDK (pip install modelvault)
• Kubernetes Helm chart for production deployment
• Model A/B testing and shadow mode
• Webhooks for job lifecycle events
• SAML/SSO for enterprise teams
• Cost and compute analytics dashboard

---

License

Copyright (c) 2026 Akhil Vase. All rights reserved.

This source code is the proprietary property of Akhil Vase.
Unauthorized copying, distribution, or modification is strictly prohibited.

---

Built for AI teams that need production-grade model ops from day one.

ModelVault -- Version it. Deploy it. Monitor it.