AI Product Engineer

Job Description:

AI Product Engineer  — Multi-Agent Workflows

About the role

We’re hiring someone to design and build AI-native, multi-agent workflows across a suite of AI-native products in cybersecurity and reliability engineering. This person will sit at the intersection of product thinking, AI systems design, workflow orchestration, and enterprise integrations.

You will help define how autonomous and semi-autonomous agents collaborate, reason, retrieve context, call tools, and take action 

This is a hands-on builder role for someone who can go from problem framing → workflow design → agent architecture → production implementation.

What you’ll do

• Design and implement AI-native multi-agent workflowsfor cybersecurity and SRE use cases
• Define how agents collaborate across tasks such as planning, retrieval, reasoning, execution, verification, escalation, and summarization
• Build agentic systems that can interact with:

• internal product capabilities
• third-party APIs
• security tools
• observability and incident systems
• knowledge bases and structured/unstructured data

• Create robust orchestration patterns for:

o   tool calling

o   state management

o   memory and context handling

o   human-in-the-loop checkpoints

o   fallback and recovery behavior

• Work closely with product, engineering, design, security SMEs, and customers to turn ambiguous workflows into production-grade AI capabilities
• Prototype rapidly, evaluate performance, and productionize successful patterns
• Design workflows for high-value use cases such as:

o   SOC alert triage and investigation

o   incident correlation and root cause analysis

o   remediation recommendations and action execution

o   runbook automation

o   cross-product enrichment and case summarization

o   integration-led automations across cybersecurity stacks

• Define quality standards for agent behavior including accuracy, explainability, observability, guardrails, and failure handling
• Contribute to the product roadmap by identifying where agentic workflows can create real user value

What we’re looking for

•       3-5+ years in product engineering, solutions architecture, workflow automation, AI applications, or related roles
• Strong experience building LLM-powered applicationsor agentic systems in production
• Experience designing multi-step workflows that combine reasoning, retrieval, API/tool use, and action execution
• Strong understanding of one or more of the following:

o   cybersecurity workflows

o   SOC operations

o   SRE / incident management / observability

o   enterprise integration platforms

• Ability to translate messy real-world operator workflows into elegant productized systems
• Experience with orchestration frameworks, agent frameworks, or custom workflow engines
• Strong coding ability in Python, Typescript, or similar
• Familiarity with APIs, event-driven systems, connectors, and systems integration
• Comfort working across product design, architecture, and implementation
• Strong written communication and systems thinking

  
Preferred qualifications

• Experience with SIEM, SOAR, EDR, vulnerability management, threat intel, ticketing, or case management products
• Experience with observability platforms, incident response tooling, cloud infrastructure, CI/CD, or platform engineering systems
• Experience building internal copilots, AI assistants, or autonomous workflows for technical users
• Familiarity with evaluation methods for LLM/agent quality, including latency, hallucination risk, task success, and safety
• Experience with RAG, memory design, planning systems, and workflow state machines
• Exposure to security and compliance requirements in enterprise environments
• Startup or zero-to-one product-building experience

Good-to-have tech platforms, frameworks, and tooling

• Hands-on experience with some of the following is a strong plus:
• LangGraph / LangChainfor stateful, graph-based agent workflows and tool-calling patterns
• ClaudeSDKfor enterprise-grade multi-agent orchestration
• CrewAIfor role-based multi-agent coordination and task delegation patterns, including production deployment options through managed runtimes  
• LlamaIndexor Haystack for retrieval-heavy workflows, knowledge-grounded agents, and document-aware reasoning patterns  
• Familiarity with MCP (Model Context Protocol)and A2A-style interoperability for connecting agents to tools, services, and other agents in a standardized way  
• Experience with model platforms such as OpenAI, Azure OpenAI, Anthropic, AWS Bedrock, or Google Vertex AI
• Familiarity with model routing, fallback strategies, latency/cost optimization, and eval-driven model selection
• Experience building with event-driven architectures, queues, webhooks, and integration middleware  
• Vector and hybrid retrieval systems such as Pinecone, Weaviate, Elasticsearch, or OpenSearch
• Experience with RAG architectures, re-ranking, knowledge stores, cache layers, and short-term/long-term agent memory
• Working knowledge of Postgres, Redis, object storage, and document processing pipelines
• OpenTelemetryfor tracing, telemetry, and workflow observability across agent/tool chains
• Familiarity with agent/LLM evaluation approaches: task success, hallucination detection, latency, cost, safety, and regression testing
• Experience with prompt/version management, replay, auditability, and failure analysis
• Understanding of guardrails, policy checks, agent governance, audit logging, and secure runtime controlsfor autonomous systems; this is becoming increasingly important as agent frameworks move into production use 

You may be a fit if you

• Think in terms of systems, workflows, and outcomes, not just prompts
• Can design when to use a single agent, when to use multiple agents, and when not to use agents at all
• Understand that production AI needs guardrails, observability, and deterministic fallbacks
• Enjoy working on hard operational domains with messy data and real consequences
• Can move fluidly between whiteboarding architecture and shipping code
• Care deeply about user trust, actionability, and measurable business impact

Success in this role looks like

• Shipping agentic workflows that users trust and adopt
• Reducing time-to-triage, time-to-resolution, and manual operational effort
• Creating reusable workflow patterns across product lines
• Balancing autonomy with control in sensitive operational environments
• Turning AI from a demo feature into a core product capability

Why join us

You’ll help shape how AI-native products are built for high-stakes technical domains. This role is ideal for someone who wants to define the future of agentic product experiences in cybersecurity and reliability engineering, while working on real customer problems and production-scale systems.