Personal project

Optitrader

A local Streamlit portfolio analytics and AI-assistant system for multi-broker portfolio tracking.

Impact

What changed

Consolidates broker data, portfolio history, daily snapshots, FX handling and AI-assisted portfolio Q&A around deterministic tools.

Role

My part

Designed and directed the product with AI-agent collaboration, focusing on practical portfolio questions rather than a polished commercial app.

Status

Active local system

The project shows how I approach finance tooling: collect messy real-world data, store it in a usable shape, and build inspection layers on top.

Essence

Project story.

Broker coverage includes Schwab live OAuth APIs and IBKR Flex data for positions, trades and cash-transfer style records.

Portfolio state is persisted locally in SQLite, with daily snapshots, position snapshots, trades and classification cache tables.

The AI assistant is designed to answer numeric portfolio questions from tools and database outputs rather than model memory.

The system also includes daily portfolio briefing/reporting work, FX conversion handling and visual exposure maps.

Technology and concepts

All projects Contact