About this role
Marvell seeks a mid-level AI Solutions Engineer in Singapore to design, deliver and support production-grade AI agents and assistants for internal business teams. The role emphasizes scoping business problems, choosing an appropriate solution and designing integrations as much as writing code.
Responsibilities
- Translate ambiguous stakeholder requests into technical designs that identify data sources, system boundaries and integration points. Decide whether to build a custom agent or configure an enterprise no-code/low-code agent-builder platform such as Glean, with future business-team ownership in mind.
- Design integrations with existing systems, REST or Graph APIs, scheduled jobs, data platforms and chat surfaces such as Slack. Plan migrations from legacy RPA approaches to API-driven pipelines, and design access controls and data permissions for mobile, web and API consumers, including models such as Databricks ACLs and Unity Catalog.
- Prototype against real data with users, refine designs, demonstrate solutions and trade-offs to stakeholders, and support adoption across teams or regions.
- Build Python backend integrations and automation, expose backend logic through API endpoints for frontend teams, and integrate LLM observability and evaluation tooling such as LangSmith to monitor groundedness, quality and cost. Document solutions, present them at architecture and compliance review boards before release, and iterate on live systems using user feedback.
Requirements
2–4 years of hands-on experience building applied AI or software solutions; strong Python and REST/Graph API skills; experience turning business requirements into solution designs; working knowledge of LLM or agent frameworks and observability tools such as LangGraph or LangSmith; familiarity with cloud infrastructure such as AWS and access-controlled data platforms; and ability to explain designs and trade-offs to non-technical stakeholders. Experience with enterprise agent-builder or search platforms such as Glean is also sought. Additional valued experience includes LLM evaluation methods, Slack API or other chat integrations, and rapid-prototyping or forward-deployed engineering.
The full-time role is eligible for bonus and equity plans, including a new-hire equity grant and annual refreshers. Applicants must be eligible to access any export-controlled information required for the position; export-license review may apply. AI tools are not permitted during interviews without prior instruction from the interviewer.