S Naga Surya AI / Backend Engineer
Resume

Case study · 2026

Apex MyDaylight

Logistics client portal and document intelligence

Classifies invoices, shipping labels and Bills of Lading, and serves the client portal’s shipment and rate-quote APIs.

Classifyinvoice, label, BoLdocument typeRouteto the right modelStructureoperational dataServeasync FastAPISecureaccess, audit logsthe pipeline I designedcustom, template or prebuilt Classifyinvoice, label, BoLdocument typeRouteto the right modelStructureoperational dataServeasync FastAPISecureaccess, audit logsthe pipeline Idesignedcustom,template orprebuilt
How Apex MyDaylight works. The coloured notes are my annotations.

What it does

Apex MyDaylight is a client portal for a logistics business moving from legacy TruckMate/.NET systems to Python services. I designed the document-intelligence pipeline that turns invoices, shipping labels and Bills of Lading into structured operational data.

I also build the asynchronous FastAPI services behind the portal’s shipment visibility, rate quotes, accounts and customer-service workflows.

How it works

  1. Classify

    Azure Document Intelligence classifies invoices, shipping labels, Bills of Lading and related logistics documents.

  2. Route

    Each document type is routed to the right model: custom extraction, fixed-template or prebuilt read.

  3. Structure

    Extracted fields become structured operational data, which also makes historical documents available to new Python workflows.

  4. Serve

    Asynchronous FastAPI services with Pydantic validation power shipment visibility, rate quotes, accounts and customer-service workflows.

  5. Secure

    Role-based access control and audit logging guard integrations with internal logistics systems, on tuned PostgreSQL and AWS.

What I built

  • Designed an Azure Document Intelligence pipeline to classify invoices, shipping labels, Bills of Lading and related logistics documents.
  • Routed each document type to custom extraction, fixed-template or prebuilt read models to produce structured operational data.
  • Build asynchronous FastAPI services with Pydantic validation for shipment visibility, rate quotes, accounts and customer-service workflows.
  • Develop React and Python modules and secure integrations with internal logistics systems, backed by role-based access control and audit logging.
  • Support modernization from legacy TruckMate/.NET systems to Python services by making historical document data available to new workflows.
  • Optimize PostgreSQL queries, indexes and connection pooling while supporting AWS-hosted components across RDS, S3 and CloudWatch.