Duties & Responsibilities
- Machine Learning & Data Analytics (Build the Business Engine)
-
• Design, build, and validate predictive models end to end — for example churn prediction, Customer Lifetime Value (LTV), and demand or behaviour forecasting — from feature scoping through to deployment.
-
• Establish A/B testing and statistical validation practices suited to a single-owner function today, documented so they can be handed off cleanly as the team grows.
-
• Partner directly with data engineering (or the equivalent function/vendor) to define the business logic, feature requirements, and data pipelines needed for model training.
-
• Own the translation of model outputs into clear, actionable recommendations for non-technical department heads and stakeholders.
-
• Track and report the business KPIs tied to deployed models (e.g., revenue lift, cost avoidance, retention improvement) — not just delivery of the model itself.
-
Generative AI Strategy & LLM Evaluation (Build the Innovation Edge)
-
• Define and prototype architecture for initial LLM/GenAI use cases (e.g., internal copilots, retrieval-augmented generation tools), designed so they can extend into multi-agent systems later.
-
• Set baseline evaluation protocols for LLM accuracy, response quality, and hallucination risk, and ensure PDPA/PII compliance before any GenAI tool goes into production use.
-
• Use AI-assisted code-generation tools to rapidly prototype and benchmark proof-of-concept GenAI solutions, and make evidence-based calls on what to scale or kill.
-
• Make and document build-vs-buy and vendor decisions (LLM providers, vector databases, cloud AI services) within an agreed budget envelope.
-
• Act as the go-to internal expert on RAG optimisation, vector database selection, and prompt engineering.
-
Foundation-Building & Scaling Readiness (The Leadership Runway)
-
• Act as the direct technical liaison to the CTO and relevant department heads, translating R&D progress and architectural trade-offs into a roadmap leadership can fund and act on.
-
• Establish documentation, coding, and model-governance standards now so the function is ready to onboard a team later.
-
• Build the business case and role specifications for the first hires into the function, in partnership with the CTO.
-
• Own responsible-AI basics for the department — PDPA/PII compliance, basic bias checks, and an incident-response approach for model or AI failures.
-
• Drive adoption of predictive and GenAI outputs among business users, so work delivered is actually used, not just shipped.