Background work with controlled retries
Used BullMQ queues to organise analysis jobs with priorities and retry handling, separating job execution from incoming API requests.
Admin · API
Automation System.
Project overview
A backend and dashboard for automated market analysis using an external AI service. BullMQ queues manage background jobs with retries and priorities, while worker threads move CPU-heavy analysis off the main event loop. Two Fastify services can be deployed and scaled independently, and a React dashboard provides visibility into bot activity, performance and analysis results.
My role
I independently built the analysis backend and React operations dashboard, including the two Fastify services, BullMQ job queues, worker-thread processing and AI-service integration.
A closer look
Choose a screen to exploreSelected contributionsEngineering
Engineering
highlights.
Technologies
Keeping heavy analysis off the request thread
Moved CPU-intensive analysis into worker threads so that processing work does not occupy the main event loop.
Services that can be deployed independently
Split the backend into two Fastify services, allowing their workloads and deployments to be managed separately.
Connecting AI analysis to a job workflow
Integrated an external AI service into market-trend evaluation and signal generation, connecting analysis requests to queued processing.
Making automated work visible
Built a dashboard for bot activity, performance metrics and analysis results, giving operators a view of background processing.
End-to-end Testing
Critical flows validated with Playwright and unit/integration coverage via Vitest.