Skip to main content
SEN

Backend & DevOps Engineer

SebastianEstradaNates

I build the systems behind the product.

Six years shipping Python backends, AWS infrastructure and data pipelines. This map shows four of those systems, and every node carries a number from production.

Get in touch
prod / 4 systems / 12 nodeslive

› saas-platform

› ec2-backups

› weather-pipeline

› data-mesh

ECS Fargate · Multi-Tenant SaaS Migration to ECS

60% faster deployments

Hover or tab through the nodes. Click one to jump to its case study.

6+yrs

building backend, cloud and data systems

60%

faster deployments after moving EC2 to ECS

50+GB

of weather data processed every day

10+TB

governed in a university Data Mesh

2h15m

backup downloads, after automating them with Lambda

GET /about

Backend engineer who ends up owning the infrastructure.

I believe in pragmatic engineering: choosing the right tool for the job, measuring everything, and building systems that are simple to understand yet powerful in capability.

I'm a Software Engineer with a postgraduate specialization in DevOps for Cloud Computing, passionate about building scalable, efficient systems that solve real-world problems. My journey in tech spans over 6 years, working across backend development, cloud infrastructure, and data engineering.

Whether it's reducing deployment times by 60%, cutting backup processes from 2 hours to 15 minutes, or processing real-time meteorological data at scale, I focus on delivering measurable impact.

My expertise lies in designing and implementing cloud-native architectures, automating infrastructure workflows, and optimizing data pipelines. I've led migrations from monolithic EC2 deployments to containerized ECS workloads, engineered microservices handling 50GB+ daily data throughput, and architected Data Mesh solutions managing over 10TB of data.

Based in
Toronto, ON, Canada
Status
Open to opportunities

GET /work

Selected work

Four production systems, in the same order as the map up top. Each one shows what changed and by how much.

› saas-platform

Multi-Tenant SaaS Migration to ECS

Architected and executed a complete migration of a multi-tenant SaaS web application from traditional EC2 instances to Amazon ECS with containerized workloads. The migration involved designing container images, implementing load balancing strategies, setting up auto-scaling policies, and ensuring zero-downtime deployment.

Deploy time

before
45 min
after
18 min
performance
60% faster deployments
scale
Multi-tenant SaaS platform
impact
Improved cost efficiency by 35%
uptime
Zero-downtime migration

› ec2-backups

Automated EC2 Backup Workflow

Designed and implemented a fully automated EC2 backup workflow using AWS Lambda, CloudWatch Events, Terraform, and Secrets Manager. The solution eliminated manual intervention, optimized backup processes, and dramatically improved recovery time objectives.

Backup download

before
2 hours
after
15 min
performance
88% reduction in backup time
automation
90% less manual intervention
reliability
99.9% backup success rate
efficiency
2 hours → 15 minutes download time

› weather-pipeline

Real-Time Meteorological Data Pipeline

Developed a comprehensive microservices architecture using FastAPI and PostgreSQL to process real-time meteorological data for watershed management. The system handles massive daily data volumes while providing low-latency access for analytics teams.

Data availability

before
6 hours
after
near real-time
scale
50GB+ daily data processing
performance
40% reduction in processing time
throughput
7 microservices handling concurrent requests
availability
Real-time data availability for analytics

› data-mesh

Enterprise Data Mesh Architecture

Designed and implemented a comprehensive Data Mesh architecture for a major university, leveraging AWS Lambda for data extraction, AWS Glue for ETL processes, S3 for data lake storage, and Lake Formation for governance. The solution provides decentralized data ownership while maintaining centralized governance.

Time-to-insight

before
baseline
after
−50%
scale
10+ TB data managed
domains
Multiple data domains with decentralized ownership
governance
Centralized governance with Lake Formation
performance
Sub-second query performance on aggregated data

Other work

4 projects

GET /stack

The stack, with receipts.

Pick a skill to see which projects and roles it actually shipped in.

Cloud & DevOps

Data Engineering

Frontend & Tools

GET /experience

Where the work happened.

Roles as spans on one timeline. Open one for what shipped.

  • Toronto, ON, Canada

    Django · React · PostgreSQL · Go · AWS · Amazon Connect

    Developing and maintaining internal web applications with Django, React, and PostgreSQL. Leading architectural improvements including migration to Go-based services.

    • Developed and maintained core features for internal web applications using Django, React, and PostgreSQL, enhancing usability and system stability
    • Prototyping Go-based replacement for Django chat service with focus on concurrency handling, service isolation, and horizontal scalability
    • Integrated frontend models with backend serializers using Django REST Framework, enabling clean separation of concerns and improved validation
    • Migrated and configured Amazon Connect contact flows and telephony routing across AWS accounts, ensuring reliable voice system operations

GET /principles

How I make engineering calls.

  • Systems Thinking

    Design for scale, resilience, and observability from day one. Every architecture decision considers failure modes, monitoring needs, and future growth.

  • Measure Everything

    Data-driven decisions through comprehensive metrics and monitoring. From deployment time reductions to processing throughput, I quantify impact to drive continuous improvement.

  • Simplicity First

    The best solution is often the simplest one that works. Avoid over-engineering while building foundations that can evolve with changing requirements.

  • Automate Relentlessly

    Manual processes are opportunities for automation. From backups to deployments, eliminating toil frees teams to focus on delivering value.

  • Infrastructure as Code

    Terraform, CloudFormation, or code - infrastructure should be versioned, reviewed, and reproducible. Configuration drift is a bug, not a feature.

  • Fail Fast, Learn Faster

    Embrace failure as a learning opportunity. Build systems with proper error handling, implement comprehensive logging, and conduct blameless post-mortems.

GET /contact

Have a system that needs an owner?

Email

Résumé

Status

Open to opportunities · Toronto, ON, Canada