AI System Architect - Fast Growing Digital Asset & Payment Company - J12415

PINPOINT ASIA INFOTECH PTE. LTD.; Singapore; On-site; SGD 100,000 - 150,000; Full time

Job summary

A leading fintech company in Singapore is seeking an experienced AI & Agent System Architect to design and deploy end-to-end AI frameworks. The ideal candidate will have over 6 years of experience in AI and ML architecture with proven skills in designing agent-based systems and strong Python expertise. Responsibilities include evaluating LLMs, integrating compliance workflows, and building scalable data pipelines. Join us to drive sustainability in the fintech space!

Qualifications

• 6+ years of experience in AI, ML architecture, or system design.
• Proven track record designing agent-based architectures.
• Hands-on experience in Python and microservices.

Responsibilities

• Design and deploy AI agent frameworks across the organization.
• Integrate LLMs and vector databases into existing products.
• Evaluate model performance and integrate compliance requirements.

Skills

AI system design

ML architecture

Python engineering

Prompt engineering

API integration

Vector databases

Tools

LangChain

OpenAI model ecosystem

Pinecone

Redis

Job description

Our client is a fast-growing digital assets and digital payment tokens company that has a global presence and holds licenses across 5 different major cities. It is headquartered in Singapore and is a pioneering green fintech company, providing regulated financial services and end-to-end carbon SaaS solutions to drive sustainability and inclusivity.

From sustainability compliance and AI-powered transportation to digital payments and wealth solutions, they deliver expertise, infrastructure, and strategy to help businesses scale new heights.

Responsibilities

AI & Agent System Architecture

• Design and deploy end-to-end AI agent frameworks across the organization, product, and operational stack.
• Build systems including autonomous agents, retrievers, action-based agents, RAG, orchestration, and API integrations.
• Architect scalable pipelines for:

1. real-time payment validation
2. compliance checks (AML/KYC/KYB)
3. client workflow automation
4. operational intelligence
5. customer service and onboarding

• Evaluate and integrate LLMs, vector databases, and retrieval systems.
• Build internal tooling, microservices, and interfaces that integrate with business infrastructure.
• Design agentic workflows to reduce manual work and enhance decision‑making.
• Evaluate model performance, hallucination safeguards, and accuracy frameworks.
• Build data pipelines, embeddings, orchestration layers, and agent‑to‑system messaging.
• Partner with Compliance, Product, and Risk teams to match operational and regulatory requirements.
• Identify automation and AI product opportunities across On/Off Ramp, Treasury, Operations, and Customer Success.
• Create frameworks and standards for AI deployment company‑wide.

Qualifications

• 6+ years’ experience in AI, ML architecture, LLM engineering, or AI system design.
• Proven experience designing agent‑based architectures (LangChain, CrewAI, Llama Index, OpenAI model ecosystem, etc.).
• Experience with vector DBs, embeddings, and orchestration tools (Pinecone, Chroma, Redis, Weaviate).
• Hands‑on engineering experience in Python, microservices, and API integration.
• Strong knowledge of prompt engineering, retrieval systems, grounding, and data pipelines.

Preferred

• Prior experience in payments, fintech, digital assets, or regulated industries.
• Familiarity with compliance workflows: KYC/KYB, AML, sanctions check.
• Experience building AI automation for B2B platforms or enterprise infrastructure.

With well over a decade of a solid and enviable track record behind us, headquartered in Hong Kong, Pinpoint Asia Infotech Pte Ltd (EA License: 16C8291) is the go‑to IT Search Firm for several top Investment Banks and Financial Institutions.

If you are interested in the above position, please send your CV to Charlie Kim [e-mail] (EA Registration number: Reg No: R23112483).