A complete application in a minute
Tailored resume and cover letter, ready to send.
Toppan Next Pte. Ltd. in Singapore seeks a versatile Full-Stack AI Engineer to drive AI transformation across diverse subsidiaries. You will bridge business and technical domains, gather requirements, prototype POCs, and deploy production AI solutions on AWS.
You will own end-to-end AI product development—from frontend Next.js interfaces to Python FastAPI backends—while establishing robust LLM workflows, MLOps, and observability to ensure measurable business impact.
We are looking for a versatile and commercially minded Full-Stack AI Engineer to drive artificial intelligence transformation across our diverse parent company portfolio. Our subsidiaries span printing & packaging, regulated communications, security, IT consulting, and language translation. Because of this multi-domain footprint, you will have a rich, continuous pipeline of unique business challenges to solve.
This role requires a true hybrid professional: part business analyst, part full-stack developer, and part AI integration specialist. You will not be siloed into just writing code or training models. Instead, you will own the end-to-end lifecycle—speaking directly with stakeholders to gather requirements, rapidly building Proof of Concepts (POCs) to validate ideas, engineering robust AI solutions, and deploying them to production on AWS. If you thrive on autonomy, enjoy full-stack development, and want to see your AI solutions generate tangible operational value across completely different industries, this is the role for you.
Act as the technical and business bridge: consult directly with internal stakeholders across our various subsidiaries to understand their workflows, pain points, and operational bottlenecks.
Gather requirements, conduct feasibility research, and architect pragmatic AI/automation solutions.
Rapidly prototype and demo Proof of Concepts (POCs) to refine product vision before committing to full-scale development.
Take successful POCs from local prototypes to fully productionized, scalable applications deployed in AWS.
Design and build intelligent agents and multi-step workflows using the Claude Agent SDK and Model Context Protocol (MCP).
Develop Retrieval-Augmented Generation (RAG) pipelines, integrating structured and unstructured data from diverse subsidiaries.
Implement function/tool calling, prompt engineering, context management, and decision logic to automate complex business processes.
Select, build, and evaluate classical ML models (e.g., classification, regression, clustering, anomaly detection) using scikit-learn where a lightweight, interpretable, or cost-efficient model is a better fit than an LLM.
Prepare and engineer features from structured business data, and validate models using appropriate metrics and cross-validation.
Combine classical ML models with LLM-based workflows (e.g., as routers, classifiers, scorers, or confidence estimators) to improve accuracy, latency, and cost.
Develop intuitive, responsive user interfaces and internal dashboards using Next.js.
Build scalable, high-performance backend services and REST APIs using FastAPI (Python).
Containerize applications and manage cloud-native deployments on AWS (e.g., ECS, EKS, Lambda, or EC2).
Establish CI/CD pipelines to ensure smooth, automated testing and delivery of software.
Implement robust tracking, evaluation, and cost-monitoring for LLM calls using Langfuse.
Set up system and application-level monitoring, alerting, and metrics visualization using Prometheus and Grafana.
Ensure AI solutions are reliable, secure, and grounded, establishing confidence thresholds for autonomous agent actions.
Establish evaluation and monitoring frameworks covering accuracy, hallucination, retrieval quality, task completion, latency, cost, and business outcomes.
Frontend: Next.js, React, Tailwind CSS
Backend: Python, FastAPI
AI & Agents: Claude Agent SDK, Model Context Protocol (MCP), LLM APIs
Classical ML: scikit-learn, pandas, NumPy
Data & Search: Vector Databases (e.g., PostgreSQL with pgvector, Qdrant, or Pinecone)
Cloud & DevOps: AWS, Docker, CI/CD (GitHub Actions / GitLab CI)
Observability & LLMOps: Langfuse, Prometheus, Grafana
Experience: 2+ years of professional experience in software engineering, applied AI, full-stack development, or DevOps
Education: Formal qualifications in Artificial Intelligence, Data Science, Computer Science, Computer Engineering—OR equivalent practical, demonstrated work experience
Builder's Mindset: Proven ability to transition fluidly between frontend interfaces, backend APIs, and cloud infrastructure
AI Fluency: Practical, hands-on experience building applied LLM applications, agentic workflows, and RAG systems (this is an applied role, not model training/research)
Classical ML Experience: Hands-on experience with classical machine learning models using scikit-learn (e.g., logistic regression, tree-based and ensemble methods, clustering), including feature engineering, model evaluation, and deploying models as part of production services
Consultative Communication: Strong business acumen and excellent communication skills. You must be comfortable leading requirements-gathering sessions with non-technical business leaders and translating their needs into technical architecture
Autonomy: High comfort level with ambiguity; you are capable of researching solutions independently and making technical stack decisions for individual projects