Back to all work
— Project 06
Backend platform

AspenServicesAPI

Internal backend platform

Aspen’s primary internal FastAPI backend for engineering data, Teams bots, Outlook add-ins, accounting reports, and PII tooling. The largest workflow is Meeting Analyzer v2, which turns raw Teams transcripts into structured meeting intelligence and verifies the output against the source before it leaves the system.

Role
Backend Engineer
Period
2023 to present
Status
Production
FastAPIAWS ECSMicrosoft GraphPresidioOpenAIDynamoDB
— Chapter 01
System shape

How the system fits together.

Click any block to read about it
Fig. 01 — AspenServicesAPI architecture
— Chapter 02
Decisions and outcomes

The calls that shaped it.

  1. 01

    Built Meeting Analyzer v2, a multi-stage NLP pipeline that turns raw Teams transcripts into structured meeting intelligence. The verifier stage cross-checks the synthesizer against the source transcript to catch hallucinations before output leaves the system.

  2. 02

    Wrote the bot-manager framework backing two Teams bots, with auto-renewing Microsoft Graph subscriptions, Presidio-based PII anonymization, and DynamoDB for session state.

  3. 03

    Shipped a capacity-testing service that runs automated performance tests against external monitoring platforms and pushes results into Microsoft Dynamics.

— Aside
The interesting work isn't the stack. It's the boundaries.
— Chapter 03
How it runs

What it runs on.

  • 01
    FastAPI single deployment on AWS ECS
  • 02
    Microsoft Graph subscriptions for Teams transcripts
  • 03
    Presidio + spaCy for PII detection and anonymization
  • 04
    OpenAI for the meeting-analyzer NLP stages
  • 05
    DynamoDB alongside Postgres for bot session state