Senior Software Engineer
VONQAug 2019 - Nov 2020
Moved core processes to Salesforce, kept job vacancy feeds running and helped build a self-organizing Scrum team, on AWS and PHP.
Technologies Used
- AWS ECS
- Terraform
- PHP
- MySQL
- Elasticsearch
- Broadway
- Datadog
- OpenAPI
- Salesforce
At VONQ I worked on two things: moving core business processes into Salesforce, and keeping the job vacancy feeds running. I also helped build a self-organizing Scrum team.
About VONQ
VONQ is a recruitment marketing company in the Netherlands. It helps employers and staffing agencies advertise jobs. A company posts a vacancy once, and VONQ pushes it to job boards, social media and niche sites, using data to recommend which channels fit that job and audience.
What I worked on
Moving to Salesforce
We migrated the core recruitment platform to Salesforce, replacing several legacy systems and the maintenance that came with them. I worked on the integration between our systems and Salesforce, mapping old data onto the new model and moving business logic into Salesforce workflows. The hard part was not the code. It was keeping sales and operations working while the ground moved under them.
Vacancy feeds
The feeds send each vacancy to many job boards in the format each one expects, thousands of vacancies a day across Western European markets. I maintained and extended them: new job board integrations, retries and error handling for when a board's API failed, and Datadog monitoring so we heard about failures before customers did.
Other work on the platform:
- Job search on Elasticsearch
- APIs documented with OpenAPI
- Webhooks for event-driven integrations
- CQRS and event sourcing with Broadway in PHP
Keeping both sides in sync
Once Salesforce held the core data, the rest of the platform still needed it. We built real-time data sync between our services and Salesforce. The Salesforce customizations themselves also needed performance tuning.
The migration also needed people to adopt the new tools. We supported that with training and documentation for the internal teams.
On the infrastructure side, services ran in Docker on AWS ECS, with EC2 and RDS underneath, all defined in Terraform and deployed through CI/CD pipelines. The feeds and their data pipelines for job listings and applicant tracking ran on a high-availability setup, with disaster recovery procedures written down.
The team
We set up a self-organizing Scrum team. That meant less about ceremonies and more about trust: people picking up work without being assigned it, and saying early when something was stuck. I also mentored junior developers.
Stack
AWS ECS, Terraform, Docker, PHP, MySQL, Elasticsearch, Broadway, Datadog, OpenAPI, Salesforce.
What I took from it
A platform migration is mostly a people project. Mapping the data was the easy half. Getting people to trust the new system was the other half.