Required Skills: Slack, PagerDuty, AKS, Azure,
Job Description
RFID is a critical complementary signal for scenarios where computer vision alone may have ambiguity, limited visibility, or difficult item-identification conditions. Accelerating RFID platform development will improve RFID-CV fusion, increase virtual-basket accuracy, and reduce member friction.
This resource will help:
- Accelerate RFID rollout to additional clubs.
- Improve RFID reader health, connectivity, coverage, and event quality.
- Build a repeatable production deployment and support model.
- Improve RFID-CV fusion for more accurate item attribution.
- Reduce ambiguous, duplicate, delayed, or missing RFID events.
- Improve monitoring, fallback behavior, and incident recovery.
- Reduce operational risk during club installation, migration, testing, and production rollout.
Scope
- Support RFID reader integration, device configuration, telemetry, and connectivity.
- Develop and maintain the RFID event-processing pipeline.
- Normalize reader events and provide stable APIs for downstream CV-RFID fusion.
- Improve handling of:
- Duplicate reads
- Missed reads
- Reader overlap
- Localization ambiguity
- Delayed events
- Reader outages
- Build reader-health, read-rate, latency, coverage, and fusion-quality dashboards.
- Implement automated alerting, recovery, and fallback.
- Support multi-environment releases from development through stage and production.
- Build repeatable deployment automation for future club rollouts.
- Create RFID test, replay, simulation, and RCA tools.
- Automate change requests and release governance through ServiceNow.
- Partner with CV, Pipeline, UX, SNG, and Ops teams to validate member-facing fusion behavior.
- Create deployment runbooks and white-glove support procedures for club onboarding.
Required skill set
- Hands-on Kubernetes experience, preferably AKS/Azure, running multi-cluster and multi-environment releases from development through production.
- Experience maintaining stable CI/CD pipelines, including:
- Rollback
- Quality gates
- Flaky-test management
- Probe-based self-healing
- Strong SRE and observability skills.
- Able to define alert rules and SLOs from logs and metrics.
- Experience routing production alerts into Slack, PagerDuty, or a comparable on-call platform.
- Experience automating change requests through the ServiceNow Change Management API.
- Strong Python development.
- Strong Kafka experience.
- Cosmos DB or comparable distributed-database experience.
- Secrets and certificate-management experience.
- Experience with event-driven and distributed systems.
- Strong API design, schema versioning, and data-contract management.
- Strong production troubleshooting and root-cause analysis.
Preferred experience
- RFID, IoT, sensor, reader, or edge-device integration.
- RFID antenna placement, reader coverage, and localization concepts.
- Low-latency streaming and event reconciliation.
- Retail or warehouse technology.
- CV-sensor fusion.
- Device-fleet management.
- Club or store deployment automation.
- Azure IoT or comparable device-management platforms.
Executive justification
These four T&M resources are not interchangeable capacity. Each closes a specific production gap:
- Model DS: Determines whether ISEE can correctly understand member actions and items.
- Item Location & Map DS: Determines whether the correct item can be associated with each action throughout the day.
- Pipeline SDE: Ensures frames, tracks, actions, and model results arrive reliably at production scale.
- RFID SDE: Adds a complementary signal that improves basket accuracy and enables scalable RFID-CV fusion.