Case studies / Expert matching system
The challenge
Faster matching, without giving up accuracy or security
The job platform needed to make search, management, and connections between experts and clients more efficient. The platform owner was also concerned about search accuracy, the security of candidate data, and system scalability. We had to design and build a system that met all 3 requirements.
What we did
Search, talk, and pay in one place
- Integrated payment system: Payments between users are completed entirely on the platform.
- Website design: We designed job posting, applications, and categories so the site can handle short- and long-term work across many fields.
- AI assistant chatbot: It helps users with questions and with finding the right work.
- Online discussion rooms: Everyone involved can meet and talk in one shared space.
Results after launch
The platform reached 29,000 sign-ups and 90,000 projects
- Reached 29,000 registered specialists and 90,000 projects: The platform grew significantly.
- Revenue up 40%: Wider use led to business results.
- More accurate search: The AI assistant greatly improved search accuracy for users.
- Fewer payment problems: The integrated payment system reduced mix-ups in transactions.
- Work moves faster: Online discussion rooms strengthened communication among everyone involved.
System architecture
The same setup in 2 zones
← Scroll sideways to see more →
Technologies used
The stack behind search and payments
| Languages | Ruby 2.1.9 |
|---|---|
| Framework | Rails 4 |
| Database | PostgreSQL, Redis |
| Search engine | Elasticsearch 5.1 |
| Batch processing | Sidekiq |
| Cloud services | AWS Elastic Load Balancing, AWS SNS |
