Middle AI Engineer

ST Engineering tuyển dụng vị trí Middle AI Engineer Tổng quan về công việc và trách nhiệm

✓ Design and build the harness layer: tool interfaces, context assembly, control loops, guardrails, human checkpoints, evaluation suites and fallback paths

✓ Build specialised agents on the platform for business unit use cases such as document extraction and reconciliation, regulatory change monitoring and BI query

✓ Build platform-level guardrails (input and output validation, PII handling, policy checks, prompt injection defence)

✓ Define fallback behaviour for low confidence, tool failure and timeout, and the human review checkpoints

✓ Run the harness test cycle on the enterprise agent platform and feed fixes back into the platform

✓ Connect agents to enterprise data sources and systems through the platform's connectors and custom tools

✓ Instrument every agent for tracing, logging, cost and latency to diagnose failures in production

✓ Configure agents against each business unit's domain rules during fit-gap delivery

✓ Review designs and code from other AI engineers and maintain the team's engineering standards

✓ Support presales with effort estimates, component diagrams and technical sections of proposals

Kỹ năng và kinh nghiệm tối thiểu

Required qualifications

✓ Bachelor's degree in Computer Science, Software Engineering or a related field

✓ 3 or more years of software engineering experience

✓ 1-2 or more years building LLM applications or agents that run in production, with evaluations and monitoring in place

Required skills and competencies

✓ Fluent English communication skills for professional, cross-functional, and client-facing working environments.

✓ Python, with production-quality testing and packaging

✓ Hands-on with an enterprise agent platform such as Dataiku, Gemini Enterprise, Vertex AI Agent Builder, Agent Development Kit (ADK) and Agent Engine, or an equivalent on another cloud

✓ Able to build the same control loop with an open framework (LangGraph or similar)

✓ Tool and function calling, structured output, retrieval, and context window management under token and cost limits

✓ Evaluation design: offline datasets, regression suites in CI, LLM-as-judge with calibration against human review

✓ Guardrail engineering: schema validation, policy checks, prompt injection defence, PII handling

✓ Observability for agents: tracing and logging with OpenTelemetry, Cloud Trace, Langfuse or similar

✓ Google Cloud (Vertex AI, Cloud Run, IAM) or equivalent, containers, and CI/CD on GitLab

✓ SQL and document parsing (PDF, email, scanned forms)

✓ Works from user stories with acceptance criteria in an Agile Scrum team

Ưu tiên có kỹ năng và kinh nghiệm

✓ Agent-to-agent interoperability experience (A2A protocol, MCP servers)

✓ Experience with a data platform (BigQuery, AWS Glue, Athena, Power BI) feeding agents or dashboards

✓ Has mentored engineers or led a small technical team

Tại sao ứng viên nên làm ở đây

ST Engineering is one of Asia's largest defense and engineering groups. It has also diversified over the years, and now supplies both military customers and commercial ones in over 100 countries, which cover its four core businesses -- aerospace, land systems, electronics and marine.

• Meal allowance & transportation allowance

• Laptop

• 100% salary from probation

• Training from probation

• Free learning of all courses on LinkedIn e-learning

• Private insurance for employees from probation

• SHUI is paid on total Gross Base salary

• Annual performance review

• Annual salary review

• • Lots of periodic company gatherings and events.

Cv gửi về: [email protected]