Solution Architecture (AI Driven Application)

About Us

Do you want to be part of Thailand banking transformation

Data is the core of the new financial services era, and we are open for the opportunity to be part to drive this change at the core.

SCB DATAx is a new venture of the Siam Commercial Bank (SCB) holdings, a leading financial services and digital services holdings in Thailand and ASEAN.

As part of the transformation of SCB into a group of product and technology companies, under the SCBx brand, SCB DATAx is the technology company to centralize data and provides AI and data science services and products to the group.

With a leading-edge cloud native data & AI platform, our vision is to support the group to providing everyone in our region with the opportunity to prosper.

We work on forward-thinking challenges of centralizing, analyzing and sharing information.

We collaborate with companies and experts in many different domains, embrace diversity and all that while having a good laugh and joy in work.

About Team & Role

At DataX, we build the technology platforms that power modern business.

Our products and services focus on cloud infrastructure, data management, and AI-driven applications.

As a member of our Enterprise Architecture team, you'll play a crucial role in connecting our technology strategy to our business vision.

Your mission is to help create and manage secure, effective solutions that drive the company forward.

You'll work alongside a talented group of enterprise, solution, and domain architects specializing in data, applications, and security.

Responsibilities

A Solution Architect who specializes in AI-Driven Applications is responsible for designing the systems that infuse software with intelligence.

They are the master planners for applications that can learn, predict, and automate.

1. End-to-End AI Solution Design

a. Architect Data & Feature Pipelines

Design the systems that take data from the data platform, clean it, and transform it into features suitable for model training. This often includes designing a feature store for consistency and reuse.

b. Design the Model Lifecycle Architecture

Create the blueprint for the entire ML workflow, including environments for data scientists to experiment, scalable infrastructure for model training, and a registry for versioning and storing trained models.

c. Select the Right AI/ML Stack

Evaluate and choose the most appropriate technologies, including ML frameworks (e.g., TensorFlow, PyTorch), cloud AI platforms (e.g., Vertex AI, Amazon SageMaker, Azure ML), and specialized algorithms.

d. Design for Inference

Architect the model deployment pattern based on the application's needs. This could be:

• i. Real-time Inference: Via a low-latency API for immediate predictions.
• ii. Batch Inference: For processing large amounts of data offline.
• iii. Edge Inference: For deploying models on devices like phones or sensors.

2. Governance, Ethics, and Security

a. Integrate principles of Responsible AI directly into the architecture. This includes designing for:

• i. Fairness: Auditing and mitigating bias in data and models.
• ii. Explainability (XAI): Implementing techniques to help understand why a model made a particular decision.
• iii. Transparency: Clearly documenting how the AI system works.

b. Ensure Security

Design the security around the AI assets, protecting models from being stolen or tampered with and securing the data used for training and inference.

c. Manage Compliance

Ensure the application and its use of data comply with relevant legal and industry regulations.

Qualifications

1. Education

a. Bachelor's Degree

A degree in Computer Science, Software Engineering, or a related field is considered the minimum requirement.

b. Master's Degree or PhD (Highly Desirable)

a postgraduate degree in a specialized field like Artificial Intelligence, Machine Learning, Data Science, or Computer Science is highly valued and often preferred.

2. Professional Experience

• a. Overall Experience: 710+ years in the technology industry.
• b. Foundation in Software/Architecture: A strong background as a software developer, cloud engineer, or a general solution architect is essential.
• c. Specialized AI/ML Experience: At least 35 years of direct, hands-on experience working on data science and machine learning projects.
• d. Production Deployment: Verifiable experience taking machine learning models from the research/prototype phase to a fully operationalized, production environment.

3. Soft Skills

• a. The ability to explain complex security topics to both technical teams and non-technical business leaders is critical.
• b. Strong analytical and critical-thinking skills to solve complex security challenges.
• c. The ability to see the big picture and design security solutions that align with long-term business objectives.
• d. Making critical decisions under pressure, especially during security incidents.

4. Technical Skills

a. Core Cloud & Software Architecture

• i. Cloud Platforms: Deep expertise in at least one major cloud provider (AWS, Azure, or Google Cloud) and its core infrastructure services (compute, storage, networking).
• ii. Modern Application Design: Strong understanding of microservices architecture, API design, event-driven systems, and containerization (Docker, Kubernetes).
• iii. Security: Solid knowledge of security principles for applications and data.

b. AI & Machine Learning Expertise

• i. ML Concepts: Deep understanding of the theory behind various machine learning models, including supervised learning (e.g., regression, classification), unsupervised learning (e.g., clustering), and deep learning (e.g., CNNs for vision, Transformers for NLP).
• ii. AI/ML Frameworks: Hands-on proficiency with Python and its core data science and ML libraries (e.g., Scikit-learn, Pandas, NumPy) and deep learning frameworks (e.g., TensorFlow, PyTorch, Keras).
• iii. Specialized Domains: Experience in at least one specific AI domain, such as Natural Language Processing (NLP), Computer Vision, Recommender Systems, or Forecasting.

c. Data & MLOps (Machine Learning Operations)

• i. Data Processing: Experience with data pipelines, feature engineering, and big data technologies (e.g., Apache Spark).
• ii. MLOps Platforms: Hands-on experience with tools and platforms that automate the ML lifecycle, such as MLflow, Kubeflow, or cloud-native platforms (Amazon SageMaker, Azure Machine Learning, Google Vertex AI).
• iii. CI/CD/CT/CM: A firm grasp of how to build automated pipelines for Continuous Integration, Continuous Delivery, Continuous Training, and Continuous Monitoring of AI models.