Agentic AI Engineer - 12 months contract

PERSOL TECH SERVICES PTE. LTD.; Singapore; Sur place; SGD 120 000 - 200 000; Plein temps

Résumé du poste

PERSOL TECH SERVICES PTE. LTD. seeks an experienced AI/ML engineer to design, build, and govern enterprise Gemini agents. You will own multi-agent architectures, integrate Gemini models with Vertex AI, and optimize grounding strategies for production use.

Role requires 7+ years in AI/GenAI platforms, strong Python skills, and hands-on experience with Gemini Enterprise, GCP, and Terraform/Kubernetes. Collaborative governance and cost-aware implementation are key.

Qualifications

• Bachelor's degree in Computer Science, IT, Software Engineering or related.
• 7+ years in AI/ML engineering, applied AI, or GenAI platform roles with hands-on agent/LLM development.
• Experience building/configuring agents on Gemini Enterprise (Agentspace) — design, grounding, data connectors.
• Hands-on with Gemini models via API or Vertex AI — prompting, tool calls, context/RAG design.
• Experience implementing AI governance controls — access mgmt, guardrails, audits.
• Understanding of LLM patterns — RAG, embeddings, vector search, multi-agent orchestration.

Responsabilités

• Design, build, and own multi-agent solutions on Gemini Enterprise, including architecture and multi-turn flows.
• Configure grounding, enterprise search, and data connectors for accurate info retrieval.
• Integrate Gemini models into agents and downstream apps; tune prompts and tool calling for reliability.
• Design and productionize RAG pipelines and retrieval strategies for grounding sources.
• Continuously evaluate agent quality with structured evaluation frameworks.
• Define and implement governance for Gemini Enterprise agents: access control, policies, workflows.
• Guard against hallucination, data leakage, prompt injection, and unauthorized data access.
• Establish monitoring, logging, and audit trails for agent behavior and policy compliance.
• Collaborate with security/legal/compliance to meet data privacy and responsible-AI needs.
• Create documentation and lifecycle standards for enterprise agents.
• Track Gemini releases and assess impact on agents/governance.

Connaissances

Gemini Enterprise

Agent development

Python

LLM prompt design

GCP familiarity

Governance & compliance

RAG & embeddings

Terraform/Kubernetes

CI/CD

Security & data privacy

Formation

Bachelor's degree in CS/IT/SE

Outils

Gemini API

Vertex AI

CI/CD pipelines

Kubernetes (GKE)

Description du poste

Duration: 1 year

Working Hours: 8.30am - 6pm, Monday - Fridays. Standby and shift work may be required

Location: Central

Key Responsibilities

• Design, build, and own multi-agent solutions on Gemini Enterprise (formerly Agentspace) — agent architecture, orchestration, task/tool design, and multi-turn conversation flows.

• Configure grounding, enterprise search, and data connectors so agents retrieve accurate, up-to-date, and properly-scoped information.

• Integrate Gemini models (via the Gemini API and/or Vertex AI) into agents and downstream applications, tuning prompts, context strategies, and tool/function calling for reliability.

• Design and productionize RAG pipelines and retrieval strategies that feed agent grounding sources.

• Continuously evaluate and iterate on agent quality — accuracy, relevance, latency, and cost — using structured evaluation frameworks.

AI Governance & Responsible Agent Operations

• Define and implement governance frameworks for Gemini Enterprise agents: access control, data permissions, usage policies, and approval workflows for new agents going into production.

• Build guardrails against hallucination, data leakage, prompt injection, and unauthorized data access across agents and connectors.

• Establish monitoring, logging, and audit trails for agent behavior, including token usage, response quality, and policy violations.

• Partner with security, legal, and compliance stakeholders to ensure agents meet data privacy, residency, and responsible‑AI requirements.

• Create and maintain documentation, review checklists, and lifecycle standards (build → evaluate → approve → monitor → retire) for enterprise agents.

• Track Gemini model and Gemini Enterprise feature releases and assess their impact on existing agents and governance policies.

Cross-Functional Collaboration & Automation

• Work closely with data science, AI engineering, security, and business teams to translate use cases into governed, production-ready agents.

• Automate agent configuration, evaluation, and deployment workflows using Python and APIs/SDKs for Gemini Enterprise and Vertex AI.

• Build internal tooling and dashboards to give stakeholders visibility into agent inventory, usage, and governance status.

• Participate in code and design reviews, contributing to shared standards for agent development and governance.

Qualifications

• Bachelor's degree in Computer Science, Information Technology, Software Engineering or a related discipline.

• 7+ years of experience in AI/ML engineering, applied AI, or GenAI platform roles, with hands‑on ownership of agent or LLM‑application development.

• Direct, hands‑on experience building and configuring agents on Gemini Enterprise (or Agentspace) — agent design, grounding, enterprise search, data connectors.

• Strong hands‑on experience with Gemini models (via Gemini API or Vertex AI) — prompting, tool/function calling, context and RAG design.

• Practical experience implementing AI governance controls — access management, guardrails, evaluation frameworks, audit logging, and responsible‑AI policies for LLM/agent systems.

• Solid understanding of LLM application patterns — RAG, embeddings, vector search, multi‑agent orchestration.

• Solid Python programming skills for automation, evaluation tooling, and API integration.

• Ability to work cross‑functionally with data science, security/compliance, and business stakeholders to govern and scale agent deployments.

• Working familiarity with Google Cloud Platform (GCP) — IAM, Cloud Storage, basic networking — sufficient to collaborate with platform/infrastructure teams.

• Google Cloud certifications (Professional Machine Learning Engineer, or Professional Cloud Architect).

• Experience with Terraform, Kubernetes (GKE), or CI/CD pipelines, for coordinating with platform/DevOps teams on agent infrastructure.

• Experience with monitoring/observability stacks (Cloud Monitoring, Prometheus, Grafana, Datadog).

• Familiarity with responsible‑AI/model‑risk frameworks applied to enterprise GenAI.

• Prior experience with other enterprise GenAI/agent platforms (e.g., OpenAI, Anthropic, open‑source LLM stacks) as a point of comparison

• Strong governance and risk mindset — able to balance agent capability with safety, compliance, and trust.

• Clear communicator who can translate technical agent behavior into business and compliance language.

• Comfortable operating in ambiguity, especially with fast‑evolving Gemini features and emerging agent governance practices.

• Ownership mentality — from agent design through deployment, evaluation, and long‑term governance.