AltsManager API

AltsManager is an open-source, privacy-oriented, AI-forward platform for tracking alternative investments — private stock, SAFE notes, real property, LLC interests, private equity, VC fund LP positions, crypto, artwork, and more — across multiple investing entities. It tracks transactions, ownership, documents (including K-1s and other tax forms), and counterparties (lawyers, accountants, fund managers), and is designed to be self-hosted so your data stays yours.

This repo is the backend API (FastAPI). It's early — currently the core ledger data model, auth and CRUD for entities, counterparties, investments, ownership, and transactions, document upload/download/review-status tracking, and a basic admin UI for browsing data. Third-party integrations (e.g. Carta) and AI agents for recurring data collection are on the roadmap.

Requirements

• Python 3.12
• Docker (for local Postgres) — Colima or Docker Desktop both work

Setup

git clone https://github.com/AltsManager/api.git
cd api
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt -r requirements-dev.txt

cp .env.example .env          # adjust if needed
docker compose up -d postgres
alembic upgrade head

# there's no signup endpoint by design — bootstrap the first admin user:
python -m scripts.create_admin --email you@example.com --name "Your Name" --password "change-me"

uvicorn app.main:app --reload

Visit:

• Swagger UI: http://127.0.0.1:8000/docs
• ReDoc: http://127.0.0.1:8000/redoc
• Health check: http://127.0.0.1:8000/health
• Admin panel: http://127.0.0.1:8000/admin (log in with the admin user above — a basic auto-generated CRUD/browse UI over the database, not the real product UI, just useful for seeing data while the API and frontend are built out)

Documents

Files are stored via a pluggable backend, set with STORAGE_BACKEND in .env:

• local (default) — kept on disk under DOCUMENT_STORAGE_PATH, nothing leaves the machine. Good fit for the "privacy-oriented" / self-hosted goal.
• s3 — any S3-compatible object store (AWS S3, MinIO, Backblaze B2, ...), configured via the S3_* settings in .env.example.

Document metadata (filename, type, tax year, review status, links to the entity/investment/transaction it belongs to) always lives in Postgres regardless of backend. Upload via POST /api/v1/documents (or the nested /api/v1/entities/{id}/documents / /api/v1/investments/{id}/documents), and download via GET /api/v1/documents/{id}/download.

Development

ruff check .                              # lint
pytest                                    # run tests (needs postgres up)
alembic revision --autogenerate -m "..."  # generate a migration after model changes
alembic upgrade head                      # apply migrations

Project structure

app/
  core/      settings, security, shared FastAPI dependencies
  db/        SQLAlchemy engine/session setup
  models/    SQLAlchemy ORM models (entity, investment, ownership, transaction, document, ...)
  schemas/   Pydantic request/response schemas
  crud/      data-access layer
  services/  document storage backends (local disk / S3-compatible) and other integrations
  routers/   FastAPI route handlers
alembic/     database migrations
tests/       pytest suite