Staff Backend Engineer

About the Company

Our client specializes in building real-time voice AI infrastructure tailored for enterprises operating in regulated industries, delivering production-scale solutions that prioritize zero downtime. They maintain robust enterprise-grade security and healthcare compliance. Having closed a seed funding round led by a prominent fund and supported by tech industry veteran investors, the company’s team remains small and senior-led, with a flat organizational structure. Reporting directly to the founder, who serves as both CEO and CTO, the engineering group is early to market and leads through demonstrated technical superiority. Current enterprise demand exceeds system capacity, necessitating this critical hire.

Key Challenges

• Determining optimal storage design for conversation records comprising audio, transcripts, traces, model invocations, and results, supporting queries ranging from single record access to multi-million aggregations.
• Defining the interface and boundary between new storage layers and existing Postgres and ClickHouse production deployments.
• Identifying first points of system degradation at peak concurrency and ensuring graceful fallback rather than total failure.
• Creating effective debugging tools for a non-deterministic processing pipeline to reduce incident diagnosis times drastically.

Responsibilities

• End-to-end ownership of the data layer, including data design, disk layout, and access optimization, with direct responsibility for incident response and pager duties.
• Improving throughput nearly tenfold while meeting stringent availability targets, proactively identifying failure modes before customers notice.
• Ensuring reliable production monitoring at scale to prevent silent data loss and guarantee event integrity.
• Reducing error investigation time from one hour to under 30 seconds in a pipeline where identical inputs can yield different outputs.
• Delivering a minimal viable version in the first week, iterating designs based on live production feedback rather than prolonged documentation phases.

Requirements

• Extensive experience in building or significantly contributing to databases, queues, or large-scale data infrastructure, able to articulate subsystem design decisions and their trade-offs, particularly in operating systems at scale.
• Proven ability to scale systems from gigabytes to terabytes daily load, managing high-cardinality data with strict service-level objectives; expert at identifying bottlenecks and quantifying the impact of optimizations.
• Deep knowledge of internals for at least one technology such as Postgres, Redis, Kafka, or ClickHouse, including behavior under distress rather than mere configuration.
• Strong familiarity with AI tools and platforms (e.g., Cursor, Claude Code, Codex), with practical insight into model selection and limitations rather than superficial usage.

Company Culture

• Small, senior-driven team with a flat hierarchy and no engineering management tiers.
• Decision-making is collaborative, involving those who implement solutions directly, without restrictive role boundaries.
• Continuous deployment several times daily with immediate fixes following customer issue reports; no separate QA division, responsibility lies with the implementer who also handles incident response.
• Rapid prototyping cycles measured in days with minimal bureaucratic overhead and no tolerance for prolonged design debates.