Platform Engineer

Founding-team role | Denver metro | Pre-seed, seed in motion SymphonIQ, Inc.We are a multi-agent AI platform built for the architecture, engineering, construction, and energy infrastructure verticals. Five core agents run in production or pilot today. We are launching with two major pilot customers right now, and real-time feedback is rolling in at a fever pitch. The seed round is in motion.We are hiring the second seat on our founding technical team, the engineer who keeps SymphonIQ alive.The RoleThis is the operator seat. You report to the CTO and own the production reality of the platform, including incident response, on-call, performance, cost, and the test harness that proves what we ship actually works. The CTO designs the system. You run it, harden it, instrument it, and tell the truth about what's working and what isn't.This is not a builder seat dressed up with a startup label. It is a real job for a real type of engineer, the operator who takes pride in MTTR, runbooks that actually get used, and a system that holds up under customer load.The MomentTwo pilot customers are live. Feedback is hot. Every day surfaces new edge cases, new performance signals, and new things the system needs to handle gracefully. This is the operator's window. Your work shapes the platform's reliability story for the seed close and beyond.Your DomainProduction operationsOn-call ownership across Bedrock, graph and retrieval layers, S3, and the agent runtime Incident response, root cause, postmortems, runbook authorship Performance and cost discipline as customer count grows, including Bedrock spend, query optimization, and capacity planning Observability through LangFuse, structured logs, metrics, and alerting that catches degradation before customers do Single-tenant isolation enforcement with IAM and networking work that holds up to enterprise scrutiny Real-time feedback loop with two active pilot customers, translating high-velocity signal into prioritized fixesVendor partner interfacePrimary technical liaison to our development vendor partners, translating priorities into delivery scope, validating what they ship, and escalating what they miss Closes the gap between vendor-delivered scope and the work that has to happen in-house Test harness Builds and maintains the automated test harness for our multi-agent platform, with end-to-end coverage that catches regressions before customers do Owns the evaluation pipeline that turns LangFuse traces into a regression suite Productionizes the evaluation methodology already documented internallyWhat we're looking for5 to 8 years building and operating production systems on AWS, with at least 2 years as the on-call lead or platform owner at a Series A or earlier-stage AI or data-intensive startup Hands-on Bedrock or equivalent managed-LLM platform experience in production, not lab work, not POCs Strong fundamentals across graph databases, OpenSearch or comparable retrieval, S3, IAM, and VPC networking Proven IaC discipline (Terraform or CDK) and CI/CD ownership Test infrastructure background, having built or substantially owned an evaluation harness for an AI, data, or distributed system Production debugging instinct. Can describe a real incident you led, what you found, what you changed, and the MTTR number that improved Comfort with multi-agent systems and the failure modes that show up when agents start interacting Composure under high-velocity pilot feedbackBonus•Prior experience as the operator alongside a strong CTO •Prior experience holding a development vendor accountable to scope and quality •LangFuse, OpenTelemetry, or comparable observability stack experience in an LLM context •AEC, construction tech, energy infrastructure, or other regulated enterprise vertical exposure •Time at AWS, an AWS Premier Partner, or a company with deep Bedrock investmentCompensationBase: $100K to $140K, pre-seed with seed close factored in Equity: 1.0 percent to 2.0 percent founding stake, four-year vest, one-year cliff The candidates who say yes to this seat are not optimizing for base. They are optimizing for ownership of a real system, direct customer contact, and equity that means something.LocationDenver metro primary. Out-of-state candidates are not prioritized at this time. US-based, work-authorized.What it isn't•Not an architect seat. The CTO owns the architecture •Not a manage-the-platform-team role. You run the system, not the org chart •Not for engineers who want to redesign everything they touch. This seat rewards the operator who makes what exists actually workHow to Reach UsDirect message Jared Smith on LinkedIn with a one-paragraph note on a real production system you have operated and the MTTR or reliability number you improved. Resumes welcome but not required up front. References, background check, and a real-time technical conversation on our actual stack are part of every conversation that moves forward.