Who you are - Have 6+ years of engineering experience, with past ownership of high-scale, production-critical infrastructure - Have experience with distributed systems and container orchestration (especially Kubernetes) - Have worked with AI/ML platforms or are excited to build foundational infrastructure for LLM-based applications - Thrive in fast-paced environments with shifting requirements and ambiguous problem spaces - Are motivated by impact, enjoy deep technical challenges, and want to work cross-functionally across security, AI, and product - Have strong familiarity with one or more parts of our tech stack: - Cloud provider: AWS - Orchestration: Kubernetes + Karpenter - LLM integration: Experience with OpenAI, Anthropic, or open-source model serving (e.g., vLLM, HuggingFace TGI, Ray Serve) - Prompt & embedding infrastructure: Vector databases (e.g., Pinecone, Weaviate, PGVector), semantic search, prompt templating systems - Datastores: Postgres + PgBouncer, Snowflake, Redis - Languages: Go and Python - Monitoring & CI/CD: Datadog, Mezmo, CloudWatch, Buildkite, CircleCI ### What the job involves - We’re looking for a software engineer to join our Infrastructure team—building and operating the core systems that power Assist, our rapidly growing AI agent platform for customer support. Assist automates support workflows across email, chat, and voice, and has grown from $0 to $1M in ARR in just 3 months. As adoption accelerates, we’re investing deeply in scaling its infrastructure to meet increasing demand and security expectations from enterprise customers - As part of the AI Infrastructure team, you’ll be responsible for the systems that enable Assist to be fast, reliable, and secure. You’ll work on foundational platform components that power real-time LLM usage at scale, while also exploring how AI can be leveraged internally to make our engineering team more productive. This team is highly cross-functional, working closely with the AI, security, and product engineering teams - This is a high-ownership role for someone who’s excited by 0-to-1 building and shaping the infrastructure backbone of our AI products - Agent service reliability and scaling: We manage and scale the infrastructure that serves LLM-powered agents across chat, email, and voice. This includes selecting inference strategies, integrating with model providers (e.g. OpenAI, Anthropic), and dynamically routing traffic for performance and cost efficiency - Prompt and embedding storage systems: Assist relies heavily on dynamically generated prompts and semantic search across support content. The team owns highly-available, fast-access storage and indexing layers optimized for real-time AI interactions - Privacy and security: Enterprises expect strict guardrails around AI use. We’re building systems like network-level intrusion detection (IDS/IPS), audit logging, and LLM usage policy enforcement to meet these expectations and unlock new sales channels - Observability and usage analytics: We operate systems that surface key metrics—token usage, latency, cost per response, and quality signals—so the Assist team can continuously improve Assist’s performance and accuracy - AI-powered developer tools: We are beginning to explore and evangelize the use of AI to accelerate internal engineering workflows—through internal chat agents, pair programming tools, and intelligent automation for deployment, debugging, and on-call. Our goal is to empower engineers across the company to build faster and more confidently with AI ### Benefits - Equity package - Insurance coverage - 401k - CSA Credit - Stipend for use at any Assembled customer - Professional development stipend - Wellness stipend