● AWS Community Day Philippines · 2026
App Development with AWS
From enterprise AI engineering to products shipped after hours.
01 · Two lives
I have two lives as a developer.
4
5
6
and many more!
02 · Ownership
Side projects are brutal engineering teachers.
When one person owns the product, every decision becomes visible.
03 · Start small
Don't architect for users you don't have.
At zero users, the simplest reliable architecture is enough.
Let real users reveal what needs to scale.
04 · Think backwards
Don't start with AWS.
Start with the problem.
4
5
6
and many more!
05 · Side projects grow up
My apps stopped behaving like side projects.
Different products. Same shift: real users arrived, and the engineering had to grow up.
06 · The moment it became real
Someone paid for it.
And everything changed.
Revenue turns a side project into a production system.
07 · Engineering becomes business
Users don't care about your stack.
They care that the product is fast, secure, and dependable.
08 · Protect what already works
Apps evolve. Regressions shouldn't.
Write tests before a change becomes a regression for 20,000 users—even if the app began as a hobby.
Tests protect the behavior users already trust.
09 · A toolbox, not a checklist
This is where AWS starts making sense.
Start with simple managed paths like Amplify and Fargate. Add services only when the product needs them.
10 · Map needs to services
Product requirement → cloud decision.
Prefer the managed service that removes the most operational work.
| Product feature | Engineering need | Possible AWS solution |
|---|---|---|
| Ship a web app | Managed frontend | AWS Amplify |
| Run an API | Managed containers | AWS Fargate |
| User accounts | Authentication | Amazon Cognito |
| Files + AI | Storage + models | S3 + Bedrock |
11 · The acceleration
Then AI changed everything.
IDEA → AI → PROTOTYPE
Hours / days12 · Engineering judgment
AI didn't delete software engineering.
PROTOTYPE ≠ PRODUCTION
13 · Apply intelligence deliberately
Not everything needs an LLM.
Deterministic software
Use it when the answer must be exact.
- Business rules
- Payments
- Permissions
AI
Use it where interpretation matters.
- Documents
- Natural language
- Reasoning
14 · Example: receipt processing
AI is part of the system. Not the system.
AI should handle intelligence. Your application should handle truth.
15 · Different constraints
Three different worlds. Same fundamentals.
Enterprise
- Reliability
- Security
- Scale
My products
- Revenue
- Users
- Retention
Hackathons
- Speed
- Simplicity
- Validation
16 · The engineering loop
The loop.
Then repeat.
17 · Right now
The barrier to building has collapsed.
One developer can now access infrastructure and intelligence that once required enormous organizations.
18 · Community
AppBuildersPH
From building my own things to finding thousands of Filipinos doing the same.
Almost 50,000developers and builders
19 · Build anyway
You don't need permission to build.
My apps started after work. Some found users, some made money, and some failed.
All of them made me a better engineer.
● App Development with AWS
Build. Ship. Learn. Scale.
BUILD. SHIP. LEARN. SCALE.
Start with a real problem. Build the simplest thing that solves it. Put it in someone's hands. Let real users expose the problems you actually have.