Agentic Engineer

About the Role

AspenView is hiring an Agentic Engineer to build, tune, and measure agentic delivery across payments and hospitality commerce for a global payments and financial-technology platform. The client is putting agents to work inside its software delivery and its banking and commerce workflows, and its AI estate is already live and large: an agentic operating system for banking built on OpenAI with Amazon Bedrock and AWS AgentCore, AI-assisted development across 8,000+ engineers, and an autonomous coding agent on core modernization. Restaurant, lodging, and hospitality commerce is a named growth vertical, so domain fluency shortens the ramp materially.

You will design and build agentic workflows, engineer the context and tool surface each agent runs on, prove results with evaluation harnesses and hard numbers, and harden everything for a regulated payments environment. We hire for demonstrated hands-on work and judgment rather than one career path — we assess the work, not the job title — and regulated-environment governance is non-negotiable.

What You Will Do

Agent Design & Context Engineering
- Design and build agentic workflows — single agents, multi-agent orchestration, and subagent decomposition — against real delivery and operational tasks.
- Engineer the context: retrieval and RAG strategy, scoping, structured summaries, and pruning rules that keep long sessions coherent and affordable.
- Own system prompts, agent instructions, and configuration as versioned, reviewed artifacts rather than ad hoc text passed between people.
- Design the tool surface each agent gets — tool inventories, MCP servers, and permission and sandboxing models — connecting agents to repositories, Azure DevOps, CI, documentation, and data platforms.
- Build on the client's existing AI estate: OpenAI APIs and the Codex toolchain, agent runtimes on Amazon Bedrock / AWS AgentCore, GitHub Copilot, and Microsoft Foundry.

Evaluation & Measurement
- Build evaluation harnesses on representative task sets — including payment, POS, and hospitality commerce flows — and instrument runs for traces, token and latency telemetry, and tool-call success.
- Track and report throughput, quality, human intervention rate, and cost per unit of work against a pre-agent baseline.

Security, Governance & Enablement
- Define the human-in-the-loop gates: what an agent may complete unattended, what needs review, and what it must never touch.
- Harden workflows against prompt injection, untrusted tool output, secret leakage, and unsafe autonomous actions, with identity-bound execution and the audit trail a regulated payments client expects.
- Codify what works into reusable configuration and playbooks, and enable other engineers to use it.

What You Bring

Education
- Bachelor's degree in Computer Science, Software Engineering, or a related technical field, or equivalent practical experience.

Experience
- Substantial hands-on delivery in a technical role — software, AI/ML, data or automation engineering, or technical leadership of engineering teams.
- Travel or hospitality domain experience is mandatory — restaurant, lodging, POS, booking, ticketing, or travel-payments systems, where folios, incremental authorisations, no-show and cancellation handling, tip adjustments, and multi-property settlement are familiar ground.
- Hands-on production work with OpenAI — APIs, the Codex toolchain, or agent frameworks built on OpenAI models.
- Production LLM or agentic systems that real users depended on: agents, agentic workflows, RAG, or LLM-assisted automation.
- Practical use of AI coding and agentic tools — Codex, Claude Code, Cursor, GitHub Copilot, or comparable — well beyond casual use.

Technical Expertise
- Python and/or TypeScript at working level, comfortable building tooling and integrations.
- Working understanding of LLM mechanics: context windows, tokenisation, retrieval, tool calling, and prompting — and where each breaks.
- Version control and CI/CD familiarity — Azure DevOps or GitHub Actions.

Soft Skills
- A measurement mindset — baseline first, instrument everything, and let the data settle trade-offs.
- English at C1 level, able to explain trade-offs to engineers and executives alike.

Nice to Have
- Payments or financial-services delivery — acquiring, issuing, core banking, or merchant processing.
- Amazon Bedrock, AWS AgentCore, Microsoft Foundry, or Azure OpenAI at production scale.
- MCP server design, tool inventories, and agent-integration surfaces.
- RAG and retrieval tuning — embeddings, vector stores, hybrid search.
- Evaluation harnesses, golden task sets, LLM-as-a-Judge.
- Token and cost optimisation: caching, compaction, model routing.
- Multi-agent orchestration frameworks — LangChain, LangGraph, or similar.
- Hospitality platform integration — PMS, kitchen display, kiosk, mobile order and pay.
- PCI DSS 4.0, SOC 2, or AI-governance and guardrail exposure.

Visa Sponsorship

AspenView does not sponsor employment visas for this role. Applicants must be permanently authorized to work in their country of residence and must not require visa sponsorship now or in the future.