Graph View FastAPI migration
Replacing legacy Django shipments graph with a configurable visualization on Orders. Aggregation, time-period, and chart-type controls share filter state with the table.
FastAPI · Lit · in-devBackend Engineer @ Grip · Remote from Colombia
Hi, I'm Alejandro Dorado, a backend engineer building production APIs for cold-chain logistics.
- Stack
- Python · FastAPI · SQL Server · Redis · Azure
- Languages
- Spanish · English · German
- Based in
- Barranquilla, Colombia · remote · hybrid in Bogotá
About
I write backend code for a cold-chain logistics platform in Miami. Perishable goods, tight delivery windows, warehouses that don't wait. That environment shapes how I think about idempotency, retries, observability, and what it means for software to truly work.
University of North Florida graduate in Computer Science, full LAC Scholarship, now shipping in production at Grip Shipping since January 2026. I ship features real users touch. A bulk upload spanning FastAPI backend and Lit frontend that saves the largest external customer ~5h every week. A billing module that captures seven fee categories the moment they're incurred instead of reconstructing them later. A webhook layer that survives retries cleanly. An observability layer that surfaces silent failures before they reach clients.
Bilingual Spanish and English, full university degree completed in English. Intermediate German. Remote from Barranquilla, Colombia, with hybrid/on-site availability from Bogotá.
LIVE Now building · in development
What I Build
01
APIs & Integrations
REST endpoints that connect external client systems to a core platform. Type-safe, documented, observable. Built for the integration layer that doesn't have a fallback.
02
Webhook Reliability
Idempotency keys, retry logic, structured error handling. The same external event arriving three times produces one record, not three. State stays consistent across distributed services.
03
Observability
Structured logging and metrics that surface silent integration failures before clients notice. Built around the principle that systems can fail in ways that look like success.
04
Data Pipelines
Bulk ingestion with row-level validation, ML feature extraction, automation scripts that surface anomalies before they escalate. From spreadsheet upload to fraud classification.
Work
-
Backend Software Engineer
Grip Shipping, Inc. · cold-chain logistics SaaS
Client Integrations team. I build backend services on Grip's proprietary Pulse OMS, the order management system that connects external client systems with the core platform.
- Bulk ASN Upload. End-to-end feature spanning a FastAPI backend (CSV parsing, row-level validation, idempotent processing) and a Lit drag-and-drop modal that surfaces errors inline. External customers now create dozens of Advance Shipment Notices from one spreadsheet. +784 LOC to production, ~5h/week saved for the largest external customer.
- Additional Fees billing module. Captures seven fee categories at the moment they're incurred (overtime labor, pallet movement, appointment delivery, late or missed appointment, reshipment, and labor reimbursement with nested audit / kitting / relabeling subtypes). Role-based access: internal ops edit, client teams view-only. Reduced billing disputes and improved reconciliation accuracy.
- Webhook + integration idempotency. Keys, retry logic, structured error handling. The same external event arriving three times produces one record, not three.
- Observability with Azure App Insights. Structured logging that surfaces silent integration failures before they reach clients.
- Authored the PRD for a dual-toggle decoupling product visibility for order recommendations vs. inventory tracking, addressing discontinued SKUs, QA holds, and seasonal items between cycles. (On the product roadmap.)
Python · FastAPI · SQL Server · Redis · DuckDB/Delta Lake · Databricks · REST · Webhooks · Azure App Insights · GitHub Actions · JavaScript · Lit
-
Programming Lab Assistant
University of North Florida
First-line troubleshooting for 900+ students per semester across lab systems, software environments, and accounts. Documented recurring issues to reduce repeat incidents and speed up handoffs to senior staff.
-
Data Analysis Intern
SalesLand Colombia SAS
Cleaned and validated recruitment and performance datasets in Python and Excel. Built automation scripts that cut manual reporting time and surfaced anomalies for leadership to act on before they escalated.
Featured Projects
Bulk ASN Upload
End-to-end feature replacing one-by-one shipment notice creation with bulk spreadsheet ingestion. FastAPI backend with CSV parsing and idempotent row-level validation, Lit frontend with drag-and-drop and inline error surfacing.
+784 LOC · ~5h/wk saved
AI JobShield
Real-time classifier identifying fraudulent job postings. Trained end-to-end on 17,000+ listings; deployed through Streamlit with explainable predictions.
FinTrack
Multi-user personal finance app. Relational schema design, JWT auth, role-based permissions, end-to-end implementation from API layer to React frontend.
PRism · PR Coverage Dashboard
- Problem. An international hospitality group's comms teams spent days manually consolidating country-level PR coverage in Excel before every executive report.
- Solution. Multi-tenant SaaS: teams upload raw coverage files, the platform validates and normalizes them into an interactive dashboard with YoY KPIs, then exports a polished PowerPoint or CSV on demand.
- Stack. Built end-to-end as sole developer — ingestion pipeline (pandas/openpyxl), ECharts dashboards, Stripe billing with webhook idempotency, JWT/RBAC auth, audit logging, bilingual UI (EN/ES). Phase 15 (production hardening), deployed on Render.
Writing
Essay · 2026 · ~5 min read
What cold-chain logistics taught me about writing software that works.
The thing about working in cold chain is that software failures have consequences you can actually see. A missed shipment notice is not just an error in a log. It is inventory arriving at a warehouse with no workflow ready to receive it. Sometimes that means product sitting on a loading dock. Sometimes it means spoilage.
Three lessons production taught me that university never did:
- Silent failures are the worst kind. Systems can fail in ways that look exactly like success: a 200 response, no exception, no alert, no trace. Observability isn't something you bolt on after something breaks. It's something you design for before it does.
- Assume every request will arrive at least twice. Idempotency keys. Check before you mutate. The fix isn't complicated; you just don't think to reach for it until you've seen what happens without it.
- Capture data at the moment it exists, not when you remember it. Reconstruction from memory is where bugs and disputes come from. Audit logs, event sourcing, write-ahead logs: same idea, every time.
School teaches you to make things work under good conditions. Production teaches you to think about what happens when conditions are not good. That is not an edge case. That is the environment your code actually lives in.
Articles
Short reflections I publish on LinkedIn, about backend reliability, what production teaches you that school doesn't, and notes from the integration layer.
See all articles on LinkedInStack
Backend & APIs
Databases
Data & Caching
Cloud & DevOps
Frontend & Other
Practices
Education
-
B.S. Computer Science
University of North Florida
Databases · Data Structures · Systems Programming · Computer Networks · Algorithms · Software Engineering. Full degree completed in English.
-
M.S. Data Science
University of North Florida · upcoming
Admitted to the M.S. in Data Science program; enrollment timing under evaluation while I focus on growing my current role.
Certifications & honors
- CS50: Introduction to Computer Science HarvardX · edX
- Back-End Developer Professional Certificate Meta · Coursera
- Back End Development & APIs freeCodeCamp
- IEEE Upsilon Pi Epsilon (ΥΠΕ) International honor society · computing & information
Languages
-
Spanish Native
-
English Advanced · Full university degree completed in English
-
German Intermediate
Contact
Talk soon?
Open to conversations about backend & integrations work, remote from Colombia, or hybrid/on-site in Bogotá.
- Email alejandro.jose.dorado01@gmail.com
- LinkedIn linkedin.com/in/alejandrodorado
- GitHub github.com/AlejandroDoradoP2002
- Open to Remote from Colombia · hybrid/on-site in Bogotá
- Timezone UTC-5 · full US Eastern overlap · Bilingual EN/ES