Fleet Reliability Engineer
Join Specter as a Fleet Reliability Engineer to enhance the reliability of our sensor fleet through data and analytics.
Posted by employer 3 months ago
First seen on Joblaze 1 hour ago
Last verified on the company career page 1 hour ago
Skills & Technologies
What you'll build
- Own the fleet’s reliability data pipeline
- Drive down observability cost
- Instrument the fleet and own health metrics
- Verify that fixes hold fleet-wide
- Build failure-mode analysis
Must have
- Strong data and software skills
- Experience operating physical or embedded device fleets
- Fluency with databases and data modeling
Nice to have
- Reliability/SRE fundamentals
- Experience across the hardware-software boundary
AI in the day-to-day
Our systems can automatically detect and reason about any physical activity captured on camera.
Not disclosed in this posting: compensation, years of experience, work arrangement, visa sponsorship.
Joblaze summary
The Fleet Reliability Engineer at Specter is responsible for ensuring the operational integrity of the company's sensor fleet by developing data-driven solutions to prevent failures. This role requires strong skills in data management, particularly with Python or Go, SQL, and observability tools like Grafana and Prometheus, as well as experience with large-scale physical device operations. Ideal candidates will have a background in reliability engineering or site reliability engineering, with a focus on building automated systems rather than performing manual tasks. Specter's small, rapidly growing team emphasizes a proactive approach to reliability in the evolving field of physical AI.
Joblaze insights
Quick facts
- What's the tech stack?
- Joblaze extracted these technologies from the posting: Datadog, Go, Grafana, OpenTelemetry, PostgreSQL, Prometheus.
- What seniority level is this role?
- Specter targets mid-level candidates for this position.
- Is this full-time or contract?
- Full-time for this Fleet Reliability Engineer role at Specter.
From the original posting
The Role
We’re hiring a Fleet Reliability Engineer to keep our sensor fleet running in the field by building the data, analytics, and recovery mechanisms that prevent failures from becoming incidents. As we scale toward thousands of sensors, fleet health becomes a data-and-systems problem.
This is the proactive, highest-leverage side of reliability: own the telemetry and data pipeline, verify that fixes hold fleet-wide, and turn field signal into cost-weighted decisions about what to fix first. Much of today’s operational load is addressable through better instrumentation, alert hygiene, and recovery verification, at little to no field cost.
Responsibilities:
Reliability Data Platform — Primary
Own the fleet’s reliability data pipeline end to end: telemetry aggregation, storage, and instrumentation.
Drive down observability cost — own the tooling spend and cut what we pay for but don’t use.
Instrument the fleet and own the health metrics that measure reliability.
Proof-of-Recovery & Alert Hygiene
Verify that fixes hold fleet-wide, not just on the device that paged.
Cut alert noise at the source — separate real failures from self-resolving ones.
Turn repeat failure patterns into automated detection and recovery.
Fleet Health & Failure-Mode Analytics
Turn fleet telemetry into a live picture of which cohorts, hardware revisions, and firmware versions are trending toward failure, and why.
Build the failure-mode analysis that tells engineering what to fix at the source.
Own fleet-wide trend and forecasting work, including power and solar planning.
Reliability Economics & Prioritization
Score reliability work in dollars — field-trip cost, hardware-return cost, observability spend — and prioritize the most expensive problems first.
Set and track the fleet’s reliability targets: uptime, offline rate, truck-rolls per sensor-year.
Give the team the data to make reliability-versus-cost tradeoffs.
Qualifications:
Strong data and software skills — Python (or Go) and SQL — and the ability to own a data pipeline end to end.
Hands-on building and tuning observability stacks (OpenTelemetry, Grafana, Prometheus, Datadog, or similar), including their cost.
Experience operating physical or embedded device fleets at scale, and reasoning about how hardware fails in the field.
Comfortable turning messy field telemetry into trends, failure modes, and forecasts.
Fluency with databases and data modeling (PostgreSQL or equivalent); infrastructure-as-code familiarity (Terraform or similar) a plus.
Bias toward building mechanisms over doing manual work.
Nice to have: reliability/SRE fundamentals (SLOs, error budgets, proof-of-recovery) applied to a physical fleet.
Nice to have: experience across the hardware-software boundary — power, connectivity, and physical failure modes.
Standard company text repeated across Specter's postings is omitted here.