The solution proposed was a single-page application that allowed the user to create questionnaires that generated instant visualizations and dashboards.


The React, Redux, Typescript combo was used to take advantage of composability and to get rid of the problems that were caused by a lack of type-safety in previous projects.

I also continued to work with Nir Kaufman’s design patterns to organize asynchronous code as message passing, ditching react-thunk in favor of middlewares.


I am growing more and more fond of Go and decided to use it for this project. I didn’t use any framework, only the following packages:

The use of migrate was great when deploying, but writing test data was a bit complicated. I should probably try Go Test Fixtures next time.


I used a Docker image deployed to ElasticBeanstalk in conjunction with a Postgres database in RDS. Configuring everything to work together proved to be quite a hassle, but fortunately you can script almost anything so next time I start a project of similar characteristics, I’ll only have to call my current setup script.

This is a small example to show you how you can achieve almost anything using the command line:

# creates a new user to 
# perform all the database-related
# operations.
aws iam create-user \
	--user-name ${DB_USER} \
	--profile iam

aws iam attach-user-policy \
--policy-arn arn:aws:iam::aws:policy/AmazonRDSFullAccess \
--policy-arn arn:aws:iam::aws:policy/AmazonVPCFullAccess \
--policy-arn arn:aws:iam::aws:policy/AmazonEC2ContainerRegistryFullAccess \
--user-name ${DB_USER} \
--profile iam

aws iam create-access-key --user-name ${DB_USER} --profile iam

read AWS_ID

aws configure set region us-east-2 --profile ${DB_USER}
aws configure set aws_access_key_id ${AWS_ID} --profile ${DB_USER} 
aws configure set aws_secret_access_key ${AWS_SECRET} --profile ${DB_USER} 

aws rds create-db-instance \
--allocated-storage 20 \
--db-instance-class db.t2.micro \
--db-instance-identifier ${DB_NAME} \
--engine postgres \
--master-username ${DB_USER} \
--enable-iam-database-authentication \
--master-user-password ${DB_PASSWORD} \
--profile ${DB_USER}

This portion of code assings AWS policies to a user to create a new RDS instance. Along with RDS, I used the following AWS services:

I used SQS to process some costly work in the background (I also tought about using RabbitMQ, but since SQS was already in AWS' ecosystem…). This mainly had to do with PDF processing. I could have triggered a goroutine but I preferred to use a separate worker.