Description

Job Description

We are placing senior AI engineers to design and deliver production agentic AI systems for an enterprise client. You will build autonomous agents on the client’s existing enterprise AI stack. Glean is the agent platform and OpenAI / GPT Enterprise provides the reasoning layer, deployed on AWS or GCP. This is hands-on build work. You own agents from design through production, monitoring, and iteration.

You approach agents as production software. You work in terms of triggers, retrieval, reasoning steps, actions, permissions, evaluation, and observability, and you know how to make an agent reliable, secure, and maintainable at enterprise scale.

What You’ll Build

• Production agents on Glean. Design and ship multi-step agents using Glean Agent Builder (triggers, steps, actions, memory), grounded in enterprise data through Glean’s connectors and retrieval.
• LLM reasoning pipelines. Configure and tune GPT / OpenAI models through Glean’s Model Hub or the direct API as the reasoning engine. Design prompts, retrieval-augmented context, and confidence and quality scoring.
• Custom actions and integrations. Build bespoke Glean Actions and integrations where no native connector exists, including auth, error handling, and rate-limit-aware batching.
• A shared agent platform. Factor common capabilities such as data access, reasoning, knowledge retrieval, and governance into reusable services so each new agent is a thin configuration rather than a fresh build.
• Evaluation and guardrails. Stand up evals, confidence thresholds, human-in-the-loop exception routing, and monitoring so agents are measurable and safe in production.

Must-Have Experience

Agentic AI and LLMs

• 3+ years building with LLMs in production systems that real users depend on.
• Hands-on experience designing agentic workflows: tool and function calling, retrieval-augmented generation (RAG), multi-step orchestration, and memory.
• OpenAI / GPT Enterprise. Deep practical experience with the OpenAI API and GPT Enterprise: model selection, prompt engineering, structured outputs, function calling, cost and latency tuning, and data-handling controls.
• Strong prompt-engineering discipline and an evaluation-first approach. You measure agent quality with evals rather than by inspection.

Glean

• Glean is required. Direct, hands-on production experience with the Glean platform: Agent Builder, Apps, Actions, Model Hub, connectors, and permission-aware retrieval. This is a non-negotiable requirement for the role.
• Working knowledge of enterprise search and RAG architecture on Glean: indexing, connectors, retrieval quality, and runtime permission enforcement.

Cloud and Engineering

• AWS and / or GCP. Production experience deploying and operating services on at least one, with comfort across both preferred. IAM, secrets management, networking, and cost awareness.
• Strong software engineering fundamentals in Python or equivalent: clean, tested, maintainable code, API design, CI/CD, and version control.
• Experience integrating enterprise SaaS systems through API (for example Jira, Slack, Google Workspace, ticketing and GRC tools), including auth flows and webhook and event patterns.
• Data engineering competence: moving and transforming data between systems, working with warehouses, and handling sensitive data securely.

Nice to Have

• Experience locking down and governing enterprise tools, including controlled intake, permission models, and audit trails.
• Familiarity with observability and incident response for AI systems: monitoring drift, alerting, and root-cause analysis.
• Exposure to regulated or finance and accounting environments where auditability and data governance matter.
• Prior consulting or client-facing delivery experience and comfort working embedded with a client’s teams

Key Skills

Education

Any Graduate