Staff AI Engineer, Agent Systems

Mollkom; Riyadh; Hybrid; SAR 240,000 - 420,000; Full time

Job summary

Mollkom is seeking an experienced engineer to lead the design and delivery of agent orchestration and tool-use systems. You will ensure durable state management, safe execution, and scalable architecture across complex workflows.

You will implement typed tool contracts, context-aware memory, and robust validation, while measuring performance and driving continual improvements in latency and cost.

Qualifications

• Experience shipping LLM and tool-use systems in production.
• Ability to evaluate action correctness and workflow outcomes beyond text quality.
• Ability to lead architecture decisions, explain trade-offs and review engineering work.
• Strong Python or TypeScript and knowledge of distributed systems, queues and retries.

Responsibilities

• Design agent and tool orchestration with durable state, interruption, resumption and human escalation.
• Build typed tool contracts, input/output validation and safeguards against duplicate side effects.
• Develop context, memory and retrieval scoped to each merchant and their permissions.
• Measure task success, latency and cost, and turn execution traces into repeatable improvement experiments.

Skills

Distributed systems

Action evaluation

Architecture decisions

Workflow outcomes

Tools

Python

TypeScript

Job description

Connect product knowledge, content, customer service and operations through clear execution paths. Lead the engineering of agents that use platform tools within understandable, testable permissions and boundaries.

Responsibilities

• Design agent and tool orchestration with durable state, interruption, resumption and human escalation.


• Build typed tool contracts, input/output validation and safeguards against duplicate side effects.


• Develop context, memory and retrieval scoped to each merchant and their permissions.


• Measure task success, latency and cost, and turn execution traces into repeatable improvement experiments.

Qualifications

• Experience shipping LLM and tool-use systems in production.


• Strong Python or TypeScript and knowledge of distributed systems, queues and retries.


• Ability to evaluate action correctness and workflow outcomes beyond text quality.


• Ability to lead architecture decisions, explain trade-offs and review engineering work.

Hybrid in Riyadh, Saudi Arabia.