TRACE Recruiter mode (30 s)
Portrait of Manuel Bravo in a black blazer, in an office

Manuel Bravo

Lead Software Engineer

I modernize critical platforms without stopping production.

Recruiter mode (30 s) See the Trace ↓

About me

I'm self-taught. I started with Delphi in 2006 and have learned every new technology since by getting my hands dirty: from VB6 and .NET to Node.js, the cloud and, now, an MBA in AI Engineering.

Away from the keyboard, I ride track days and play video games. Both teach what the job does: measure before you touch anything, get the details right, and improve one lap at a time.

I want to keep growing as a technical leader, keeping up with how software engineering changes and mentoring the developers on my team.

How I work

  • Hands-on Code, pull request reviews and architecture decisions, together with the team.
  • Quality over quantity Automated tests and quality gates before anything reaches production.
  • Simple first The simplest solution that solves the real problem, without over-engineering.
  • Active listening Understand what the business and the stakeholders need before proposing a solution.
  • Focus on what I can change I separate what I control, what I can influence and what is only a concern. The energy goes to the first two; the rest I keep an eye on without letting it stall the team.
  • Mentoring Help every developer on the team grow, with code reviews that teach.

The Trace

Follow one request from the phone to the database. Each stop is a layer I modernized, with the measured result.

Stage 1

Client

A 2016 Ionic app where a critical booking journey took 12–20 s.

12–20 s ~1.2 s~10–16× fasterAppointment-booking journey · Porto

My manager asked for a benchmark: the Ionic app against native. I ran it and presented the numbers. This simulation replays it.

Benchmark: booking journey

Legacy · Ionic (2016): 12–20 s

Native · Swift/Kotlin: ~1.2 s

~10–16× faster

Speed is only what a benchmark measures. Going native also put the app closer to the hardware, opening the door to wearables.

Simulation based on measured production timings. No real systems are called.

Open case
Problem
A 2016 Ionic app where a critical booking journey took 12–20 s.
Action
Led the migration to native iOS (Swift) and Android (Kotlin); moved regression testing to Compose Testing + XCUITest.
Result
Booking ~1.2 s.
  • Swift
  • Kotlin
  • Jetpack Compose Testing
  • XCUITest
  • WPF

Stage 2

Edge & Security

Fraud in reimbursement workflows.

baseline −40 %Reimbursement fraud · Porto

Open case
Problem
Fraud in reimbursement workflows.
Action
Added facial biometrics to reimbursement.
Result
Fraud −40 %.
  • Biometrics
  • AWS
  • Firewall troubleshooting

Stage 3

API Gateway

High platform cost on CA API Gateway.

CA API Gateway Sensedia−30 %API platform cost · Porto

Open case
Problem
High platform cost on CA API Gateway.
Action
Migrated the gateway to Sensedia.
Result
Cost −30 %.
  • Sensedia
  • CA API Gateway
  • REST
  • SOAP

Stage 4

Services

Legacy .NET/SOAP services with a large vulnerability surface.

legacy .NET/SOAP Node.js (~80 % migrated)−70 %Security vulnerabilities · Porto

Open case
Problem
Legacy .NET/SOAP services with a large vulnerability surface.
Action
Modernized ~80 % of them to Node.js.
Result
Vulnerabilities −70 %.
  • Node.js
  • TypeScript
  • C#
  • .NET
  • Microservices

Stage 5

Cache & Messaging

Intermittent failures and duplicate processing.

8 h 4 h−50 %Critical processing workflow · Porto

Open case
Problem
Intermittent failures and duplicate processing.
Action
Message idempotency + Redis caching; reworked the architecture of a critical workflow.
Result
Workflow 8 h → 4 h; better reliability.
  • Redis
  • Idempotency
  • AWS

Stage 6

Database

Deadlocks and slow batches on high-volume data.

8 h 2 h−75 %Critical batch (rental platform) · Nunes Imóveis

Open case
Problem
Deadlocks and slow batches on high-volume data.
Action
Tuned queries, transactions, locking strategy and stored procedures.
Result
Batch 8 h → 2 h; deadlocks eliminated; +20 % throughput.
  • SQL Server
  • Oracle
  • MySQL
  • LINQ

Stage 7

Observability & Quality

Production issues that span the app, chatbot and web; releases without a safety net.

baseline +40 %Developer productivity (AI pilot) · Porto

Open case
Problem
Production issues that span the app, chatbot and web; releases without a safety net.
Action
End-to-end RCA with OpenSearch; Selenium / Postman / JMeter; BDD/TDD; 90 % Sonar gate; Copilot + Claude pilot.
Result
+40 % productivity; load tests before go-live.
  • OpenSearch
  • Selenium
  • JMeter
  • Postman
  • Sonar
  • GitHub Copilot
  • Claude

Journey

Where I've worked and what changed at each stop.

  1. 2022 → now Porto · Lead Engineer Porto Saúde digital modernization (Trace stages 1–7)
  2. 2022 Digisystem · Operation Manager Pre-sales and solution design for AMS / Service Desk / Cloud across BR + LATAM
  3. 2019 → 2022 Indra · Project Manager Waterfall → Scrum, +30 % delivery speed; ~32 % margin sustained
  4. 2018 → 2019 Indra · Lead Software Engineer Overtime 40 → 10 h/mo; RFPs worth R$500K–1M
  5. 2017 → 2018 Indra · Senior Software Engineer SLA 70 % → 98 % (ITIL / HPSM)
  6. 2016 → 2017 Linx · Enterprise Software Engineer WPF POS +15 % responsiveness
  7. 2013 → 2016 COREON · Software Engineer +20 % performance for financial/insurance clients
  8. 2009 → 2013 DIMEP · Software Engineer VB6 → C#/WPF; TCP/IP hardware SDK; device cost −73 %
  9. 2008 → 2009 Nunes Imóveis · Software Engineer VB6 / Classic ASP → .NET; batch 8 h → 2 h
  10. 2006 → 2008 Early career · Software Developer Delphi · Pascal

Toolbox

Tools grouped by area. No percentage bars: they don't mean anything.

Languages & stacks
  • C#
  • .NET / .NET Framework
  • ASP.NET
  • Node.js
  • TypeScript
  • Angular
  • Java / Spring Boot
  • WPF
  • Delphi
  • Pascal
Testing & quality
  • Selenium
  • JMeter
  • Postman
  • BDD/TDD (Given-When-Then)
  • Sonar 90 % gates
  • Jetpack Compose Testing
  • XCUITest
APIs, cloud & architecture
  • REST
  • SOAP
  • Microservices
  • AWS
  • CA API Gateway
  • Sensedia
  • Redis
  • Message idempotency
Observability tools
  • OpenSearch
  • End-to-end root-cause analysis
Databases
  • SQL Server
  • Oracle
  • MySQL
AI-assisted coding
  • GitHub Copilot
  • Claude
Frameworks & methodologies
  • Scrum
  • ITIL / HPSM
  • Pre-sales & RFPs
  • Budgeting
  • Team building

Garage

What I do when I'm not at the keyboard, and what it taught me about the work.

See all photos and videos →

Portrait of Manuel Bravo in a black blazer, in an office

Contact

Looking for someone to modernize a critical platform without stopping production? Let's talk.

Talk on LinkedIn ↗Download CV (PDF)

Contact via LinkedIn only

An unhandled error has occurred. Reload 🗙