AI Systems Engineer

Qlik•Boston, MA

1d•$100,000 - $125,000•Hybrid

About The Position

As an AI Systems Engineer on Qlik’s AI Practice team, you will own the infrastructure, deployment, and reliability of production-grade agentic AI systems that power the next generation of intelligent automation at Qlik. Reporting to the Global Head of AI Strategy, Policy, and Governance, you will operate at the intersection of cloud infrastructure engineering and AI systems, provisioning, securing, and scaling the AWS environments that make our AI pipelines run reliably and cost-efficiently.

Requirements

Strong hands-on Terraform experience. You version-control infrastructure the same way you version-control code.
Deep AWS experience across Bedrock (model invocation, agents), Lambda, API Gateway, ECS/Fargate, App Runner, IAM, VPC, S3, CloudWatch, and X-Ray.
You understand the AWS IAM model deeply: roles, policies, SCPs, permission boundaries, and cross-account trust. You instinctively scope to least privilege.
You build pipelines that deploy infrastructure and application code reliably. GitHub Actions, CodePipeline, or equivalent. You know how to roll back safely.
Experience with OpenTelemetry, CloudWatch Logs Insights, and distributed tracing. You care about visibility into what AI systems are actually doing and what they cost.
Comfortable with Python for scripting, Lambda functions, and lightweight automation. You don’t need to be an ML engineer, but you can read and modify agent code.
Comfortable with containerized workloads (Docker, ECS, Fargate, App Runner).
You understand vector search infrastructure and are comfortable operating OpenSearch clusters for semantic retrieval.
You close the loop. You don’t just provision infrastructure, you monitor it, cost-optimize it, and improve it without being asked.
Energized by ambiguity and comfortable writing the runbook that didn’t exist before you joined.
Able to explain an IAM boundary decision or a rate-limiting architecture to an AI engineer or a business stakeholder equally well.
You think about what happens when things fail, when costs spike, when an agent calls an endpoint it shouldn’t. You build guardrails proactively.

Nice To Haves

Bedrock AgentCore exposure is a strong plus.
Experience running long-running agent processes or streaming inference endpoints is a plus.
Experience with RAG pipeline data stores preferred.

Responsibilities

Design and maintain infrastructure-as-code (Terraform) for all AI workloads: Bedrock agents, AgentCore runtimes, Lambda functions, API Gateway endpoints, and supporting data services.
Own deployment, configuration, and scaling of AWS Bedrock AgentCore environments, including model invocation routing, session management, and memory backends.
Architect secure, least-privilege IAM policies for agent runtimes, MCP integrations, cross-account access patterns, and service-to-service authentication.
Build and maintain CI/CD pipelines (GitHub Actions, AWS CodePipeline) for agentic workloads, covering infrastructure changes, Lambda deployments, and agent configuration updates.
Instrument AI systems with OpenTelemetry, CloudWatch, and X-Ray. Own observability strategy: traces, metrics, cost dashboards, and alerting across agent pipelines.
Manage networking topology for AI workloads: VPC design, PrivateLink, security groups, and egress controls to ensure data never leaves governed boundaries unexpectedly.
Build and operate the data infrastructure supporting agents: OpenSearch for semantic search and vector retrieval, and S3 lifecycle policies for knowledge artifact storage.
Collaborate with AI engineers on MCP server deployments, containerized agent runtimes (ECS/Fargate), and performance tuning to hit latency and cost targets.