Andrew DoPlatform delivery
Independent systems

Small tools built around recurring problems.

These builds show how I frame a problem, choose an appropriately small operating model and connect existing services without creating unnecessary product surface.

Personal build

Portfolio Monitoring Assistant

A scheduled research workflow that combines portfolio data and relevant market information into a structured Telegram briefing.

Problem

Reviewing holdings, market movement and relevant news manually created repeated work without a consistent decision record.

Design & Approach

Designed a scheduled pipeline that gathers inputs, structures analysis and sends a concise briefing to Telegram.

Result

Reduced the manual steps required to prepare a repeatable personal portfolio briefing.

Execution Flow

Portfolio Monitoring Assistant pipeline flow

  1. Portfolio data — Holdings list (Andrew owned)
  2. Market and company information — External financial APIs (Partner-owned delivery)
  3. Scheduled analysis pipeline — Python parsing service (Andrew owned)
  4. Structured decision context — Information filters (Andrew owned)
  5. Telegram briefing — Private bot notification (External approval gate)

Product Decisions

  • Use an existing daily channel rather than create another app.
  • Schedule execution instead of continuously polling.
  • Keep data, analysis and message formatting separate.
  • Keep the workflow personal research, not financial advice.

Technical shape

Python pipeline servicesCron-scheduled task executionTelegram Bot API integration