Senior AI Engineer

Role overview

Senior AI infrastructure engineer building the services that connect enterprise applications and systems with large language model providers. The focus is on production-grade gateways and supporting services for routing, authentication, authorization, rate limiting, protocol translation, tenant isolation, streaming, and observability. This is a systems-oriented role for someone comfortable owning services from design and implementation through deployment, monitoring, and continuous improvement.

Responsibilities

- Design and develop MCP and AI gateways that mediate between applications, agents, and model providers. - Implement protocol-level capabilities including MCP clients and servers, request routing, streaming, tool-call proxying, authentication, authorization, and tenant isolation. - Build high-throughput, low-latency services involving TCP, TLS, HTTP/1.1, HTTP/2, JSON streaming, connection pooling, and backpressure. - Own application-level PostgreSQL concerns such as schema design, indexing, query planning, transactions, connection management, and large-scale migrations. - Design concurrent systems using worker pools, queues, graceful shutdown, and race-free shared state; profile services to identify and resolve bottlenecks. - Establish useful observability for AI workloads, including metrics, tracing, logs, token usage, provider latency, cache performance, failure modes, and request cost.

Requirements

- At least five years of senior-level experience shipping and operating production services in Go, Ruby, or both. - For Go-focused candidates, strong knowledge of the standard library, networking, TLS, concurrency, memory behavior, allocation patterns, garbage collection, and lock contention. - For Ruby-focused candidates, substantial Rails production experience, including ActiveRecord tuning and performance under load. - Deep application-side PostgreSQL knowledge covering optimization, indexing, transactions, pooling, and migrations. - Practical experience with concurrency, profiling, Kubernetes, containers, deployment pipelines, security, and distributed-system troubleshooting. - Experience integrating with LLMs at the protocol level, including message formats, tool calling, streaming, caching, provider APIs, token usage, latency, and cost.

Nice to have

- Experience building or integrating MCP servers or clients and understanding transport and capability negotiation. - Ability to audit and verify model output, identify security issues or hallucinated interfaces, and explain the underlying technical cause. - Interest in evaluating AI-assisted development workflows, automated code review, agent tooling, and internal developer systems.

Benefits and work setup

- Flexible, trust-oriented working culture with significant ownership and emphasis on sustainable productivity. - Employees may use AI and LLM development tools according to their preferences, with access to a broad range of models and coding tools. - The source describes a benefits offering intended to support employees both at work and outside it, without specifying individual benefits.