Location : Hyderabad, India
About the role : Build a compact Complex Event Processing engine that matches multi-actor sequences across cameras and sensors with out-of-order events and watermarks. Think “playbook DSL” + fast state machines.
What you’ll do
- Implement sliding / tumbling windows, joins, late-event handling, and watermarks.
- Design a small pattern DSL : LOOKOUT@Door → LOOKOUT@Wall → THROW@Fence.
- Maintain a zone graph + path feasibility to constrain matches.
- Ship production-grade async Python (FastAPI / uvicorn) with profiling and tests.
Must-have
Strong data-structures & systems thinking; async Python; complexity analysis.Streaming mindset : throughput vs latency, back-pressure, batching.Comfort turning specs into clean, tested code.Nice-to-have
Go for high-throughput services; Protobuf / LZ4; MQTT / Service Bus.Any exposure to DeepStream / GStreamer / graph analytics.Impact
Your matcher is the brain that turns detections into actionable incidents with evidence.Next steps : If interested please email your cover letter and resume to
Ashwin at ashwin.jaiswal@safespaceglobal.ai