Senior Staff Software Engineer, Recognition Platform

Metropolis•New York, NY

3d•$241,000 - $279,500•Onsite

About The Position

Metropolis is seeking a Senior Staff Software Engineer to serve as the technical owner of the Metropolis Recognition Platform (MRP). The role leads the platform’s architecture and evolution across application engineering, computer vision, data, security, platform teams, and external partners. The engineer sets technical direction, writes architectural RFCs, and drives cross-functional alignment without formal reporting authority. The position focuses on high-volume edge-to-cloud pipelines and decisioning systems that turn physical-world sensor signals into real-time actions. It includes overseeing architecture from edge ingestion through identity resolution, decisioning, notification, and operator surfaces; designing privacy, identity, and consent models for biometric-regulated environments; and owning production quality through service-level objectives, alerting, runbooks, testing, and progressive rollouts. The role is part of an office-first corporate team. Employees are required to work on-site at least four days per week. The anticipated base salary is $241,000–$279,500 USD annually. Total compensation may also include healthcare benefits, a 401(k) plan, disability coverage, basic life insurance, stock options, bonus plans, and other benefits.

Requirements

12+ years of experience building and operating distributed, data-intensive production systems, including deep experience as a multi-service platform technical owner.
Proven track record of cross-functional technical leadership and aligning independent engineering teams on shared architecture.
Expert proficiency in a JVM language such as Java or Scala, with fluency in event-driven and service-oriented design using gRPC/protobuf, streaming, and relational data modeling.
Hands-on experience designing high-throughput ingestion pipelines and deep understanding of delivery semantics, idempotency, ordering, backpressure, and tail latency.
Experience building rules, eligibility, or decisioning systems where deterministic logic coexists with probabilistic or model-driven signals.
Strong systems judgment integrating third-party APIs, SDKs, and vendor constraints.
Excellent written technical communication skills and a track record of writing decisive RFCs.
Comfort navigating physical-world deployment constraints, hardware conditions, and operational messiness beyond staging environments.
Deep understanding of observability, automated testing, and progressive delivery.
Fluency with AI-assisted engineering tools such as Claude Code and GitHub Copilot in daily workflows.

Nice To Haves

Experience with computer vision, biometric matching, or sensor-fusion systems, including reasoning about confidence, thresholds, and drift.
Background operating in privacy- or biometric-regulated domains such as BIPA, CCPA, or GDPR, or working with consent-management systems.
Experience with IoT, edge computing, fleet-scale device telemetry, or MQTT-class transports.
Expertise in identity resolution or customer data platforms, stitching multi-system signals into durable profiles.
Experience with hospitality, retail, or physical-operations software, including hospitality CRM and PMS integrations.
Prior experience in a Uber-tech-lead, TL-of-TLs, or principal-adjacent scope.

Responsibilities

Lead end-to-end architecture across edge ingestion, identity resolution, decisioning, notification, and operator surfaces; drive spec-driven development through RFCs and manage roadmap sequencing.
Drive cross-functional alignment and resolve ownership ambiguities across application, computer vision/machine learning, data, security, and platform engineering teams, as well as external partner integrations.
Evolve high-volume edge-to-cloud streaming pipelines on a horizontally scalable backbone, engineering for tail latency, intermittent connectivity, backpressure, and out-of-order events.
Design rules and decisioning layers for real-time actions, including watchlist alerting, arrival routing, eligibility, consent gating, frequency capping, and next-best-action logic.
Balance deterministic logic with probabilistic signals by establishing explicit thresholds, quantitative cost reasoning for recognition errors, and replay harnesses for model testing.
Architect privacy, identity, and consent models for biometric-regulated environments, including bounded-time opt-out propagation, in collaboration with Legal and Security.
Own production quality by defining SLOs, actionable alerting, runbooks, soak testing, and progressive rollout via feature flags.
Elevate technical standards through design reviews, code reviews, mentorship, and AI-native engineering workflows.