Applied AI Engineer

PebbleRoad is a leading Singapore-based AI services company dedicated to orchestrating intelligent, end-to-end business workflows for large organisations across financial services, government, healthcare, and more. With over 20 years of experience, we move organisations past operational bottlenecks by building AI-native processes that transform complex, unstructured documents into reliable business data. Our approach, centred around services, solutions, and products like DocuPrism, delivers dramatic improvements in efficiency, governance, and speed across core business operations.

We are now strategically expanding our team to scale our AI services and accelerate the development of our AI products pipeline. Join us in shaping the future of intelligent automation.

The role

We are looking for an Applied AI Engineer to build practical AI capabilities for our product, DocuPrism, and our client solutions. The role focuses on the AI layer: designing AI workflows, selecting suitable models, evaluating output quality, and exposing AI features through APIs for our backend, OutSystems and Camunda developers to consume.

You will work closely with product managers, engineers and clients to turn business problems into reliable AI-powered features.

What you'll do

Design and develop AI services for real product use cases, mainly document intelligence for insurance and healthcare

Build APIs for AI capabilities so they can be integrated into existing systems and workflows

Develop OCR and document intelligence solutions that cope with varying layouts

Use LLMs for extraction, classification, summarisation and decision support, with structured outputs

Evaluate AI outputs for accuracy, relevance, consistency, and reliability

Create fallback logic, confidence scoring, and guardrails for uncertain AI outputs

Monitor AI performance and improve models or prompts over time

Join client sessions to scope and demo AI features

What we're looking for

2+ years of software or ML engineering, with AI features that have gone beyond proof-of-concept into production

Strong experience in AWS stacks (e.g. Bedrock, Textract, Lambda, SageMaker)

Experience working with LLMs, prompt engineering, structured outputs, and AI APIs

Experience with OCR/document AI, especially extracting data from forms, claims, invoices or receipts

Understanding of embeddings, similarity matching and RAG, with tools like LangChain, LlamaIndex or Hugging Face

Ability to evaluate AI quality using test cases, benchmarks, and error analysis

Basic understanding of cloud deployment, Docker, CI/CD, logging, and production reliability

Nice to have: insurance or healthcare domain experience, Camunda or OutSystems exposure

Location and work setup

We're based in Singapore, with our office at 52a Tanjong Pagar Road. We're remote-friendly: come in for team days, client sessions and celebrations, and work where you work best the rest of the time.

Benefits

Competitive compensation with performance-based bonuses

Medical insurance

Flexible, remote-friendly work

Real, enterprise-grade AI systems, not demos

Access to training and learning resources to deepen your AI and cloud expertise

PebbleRoad is a strategy, design and innovation practice. We help organisations thrive in a digital age by aligning strategic vision, digital experiences and organisational capabilities.

We started our practice in 2004 with one objective: do impactful work. This single-minded pursuit has guided who we hire, who we work for, the work we do and the decisions we make.

We worked with clients such as DBS (Singapore), Digi (Malaysia), Sinar Mas (Indonesia), ADB (Philippines), CLP (Hong Kong) and many government agencies in Singapore.