Forward Deployed Engineer

What is a Forward Deployed Engineer at Sherlocks AI?

A Forward Deployed Engineer at Sherlocks AI is a customer-embedded engineer who owns the technical outcome of a deployment.

You sit inside the customer's engineering org during deployment, understand their production environment as if it were your own, and make Sherlocks AI work for their specific stack.

Palantir invented this role. OpenAI, Anthropic, Ramp, and Databricks all hire for it because the gap between an AI product working in a demo and working in a customer's production environment is enormous. That gap is where you live.

Responsibilities

• Embed with customer SRE, platform, or DevOps teams during the first two to six weeks of deployment. Learn their observability stack, alerting patterns, runbook culture, and incident response process.
• Deploy Sherlocks AI into their environment. Configure agents against their specific integrations (Datadog, New Relic, Grafana, Prometheus, ELK, PagerDuty, Slack, Kubernetes, cloud provider APIs).
• Debug integration issues live on production data without breaking anything.
• Investigate real incidents alongside customer on-call engineers. Note where Sherlocks AI works, where it misses, where it needs additional context.
• Own the operational performance and emergency response for Sherlocks AI in customer environments. When Sherlocks AI breaks on customer prod at 3 AM, you are on the bridge.
• Contribute production code, integrations, and deployment scripts directly into Sherlocks AI and, when needed, into customer codebases.
• Identify and codify repeatable deployment patterns. Contribute insights back to our Product and Engineering teams weekly.
• Write internal deployment guides so the next customer with a similar stack ships faster.
• Travel to customer sites when it accelerates deployment. Roughly 20 to 30 percent travel expected.

You may be a good fit if you have

• 4+ years of production engineering experience. SRE, platform engineering, or infrastructure engineering background preferred. You have been on-call. You have written a postmortem. You have felt real production pain.
• Strong Python and at least one systems language (Go, Rust, or similar).
• Deep working knowledge of at least one modern observability stack (Datadog, New Relic, Grafana + Prometheus, or similar) and at least one cloud provider (AWS, GCP, or Azure).
• Kubernetes fluency. You do not need to have written a controller, but you should be able to debug a CrashLoopBackOff without Stack Overflow.
• Customer-facing communication skills that a technical audience trusts. You can explain a technical tradeoff to a Fortune 500 SRE lead without oversimplifying or overhyping.
• Comfort with ambiguity. Customer deployments will surface problems Sherlocks AI has never seen. You will not have a playbook. You will be the one writing it.

Why Sherlocks AI

• High growth. The customer base and the product surface are both expanding faster than the team, which is why this role exists.
• Small team. There is no layer between you and the people building the product, and no queue to get a decision made.
• High impact. The deployments you own decide whether enterprise customers succeed, and what you learn in the field sets the product roadmap.
• Large market. Every company running production infrastructure has the incident response problem this product solves.
• Category leadership. AI SRE is being created right now, and the deployment patterns you establish here are the ones the rest of the market will copy.