Xionex Hiring! Full Time Agentic AI Engineer in Federal Territory - Ricebowl

About XIONEX

XIONEX is building X-One, an AI-native banking operations platform combining agentic AI, deterministic controls, human oversight and auditable workflows.

XIONEX also operates as an AI-native company: agents perform the work they can perform; humans remain accountable for review, judgement and final decisions.

We intend to apply this model not only within X-One, but also across our internal operations, including finance, HR, procurement, marketing, sales and other recurring business processes.

The Role

You will design and implement AI and agentic capabilities primarily for X-One, including orchestration, retrieval, tool use, evaluation, model integration and human-review mechanisms.

You will also help XIONEX identify and automate suitable internal processes. This includes building controlled agents and workflows that support everyday work without removing human accountability.

This is not a prompt-only role. It requires strong engineering foundations and the ability to turn AI concepts into reliable operational systems.

Candidates with strong backend, data or applied AI experience who have not yet specialised deeply in agentic AI are encouraged to apply, provided they are motivated to learn quickly. Practical agentic AI experience remains a strong advantage.

What You Will Do

• Design and implement agentic workflows and multi-step AI pipelines for X-One.
• Build RAG, document extraction and structured-output capabilities.
• Integrate LLMs with APIs, databases, rules engines and business workflows.
• Develop internal agents supporting functions such as finance, HR, procurement, marketing and sales.
• Analyse processes and determine which tasks can be delegated safely to agents.
• Define safe tool-calling patterns, permissions and execution boundaries.
• Implement human-in-the-loop review, approval and escalation mechanisms.
• Establish evaluation frameworks for quality, drift, hallucination, latency and cost.
• Version and trace models, prompts, tools, configurations and decisions.
• Design fallback, rollback, kill-switch and failure-handling mechanisms.
• Build observability for AI decisions and agent activity.
• Work with the Software & Systems Architect to separate probabilistic AI behaviour from deterministic business logic.
• Contribute directly to Python and FastAPI services and production debugging.

What You Bring

• Strong Python and backend engineering experience.
• Practical experience integrating LLMs, machine learning or AI services into applications.
• Experience with RAG, vector databases, semantic search or document intelligence.
• Understanding of APIs, data flows and workflow automation.
• Strong interest in agent orchestration, tool calling and controlled autonomy.
• Ability to create automated evaluations and regression tests for AI systems.
• Understanding of model limitations, failure modes and safe fallbacks.
• Ability to debug the complete AI pipeline, not only prompts.
• Strong ownership, analytical thinking and clear communication.
• Willingness to work in a human-agent cooperation model.

Useful Experience

• FastAPI, PostgreSQL, pgvector and asynchronous Python.
• LangGraph, LangChain, LlamaIndex or comparable orchestration frameworks.
• Multi-agent systems or model-agnostic architectures.
• Prompt and model versioning, tracing and observability.
• Business-process automation or internal enterprise tooling.
• Banking, fintech, compliance or another regulated environment.
• Human oversight, explainable AI or controlled decision automation.
• OCR, document processing or structured financial data extraction.

What Success Looks Like

Within the first year, XIONEX has:

• reliable and measurable agentic workflows in X-One;
• controlled model and tool execution;
• systematic evaluation and regression testing;
• complete traceability of AI-supported decisions;
• reusable agents supporting selected internal business processes;
• and a scalable human-agent cooperation model across the company.

What We Offer

• Permanent full-time employment.
• Hybrid working in Kuala Lumpur.
• Direct collaboration with founders, architects and banking-domain specialists.
• 13% employer EPF contribution.
• Medical benefits and role-dependent annual leave.
• The opportunity to build applied agentic AI for both regulated banking operations and the AI-native company behind them.