New opportunity

Software Development Engineer in Test

Plaud · Singapore

About this role

Plaud is seeking a Software Development Engineer in Test to join their Global Product R&D Center in Singapore. This role focuses on ensuring the quality and reliability of AI-driven products, including recommendation systems, AI agents, and core server infrastructure. Responsibilities include: Server-Side & Architecture QA: Executing functional, performance, and stability testing for core modules and leading API regression and data-consistency checks during architecture upgrades. AI System & Algorithm Evaluation: Designing quality gates for retrieval and memory systems, long-tail recall, and embedding drift for recommendation/search systems. AI Infra & Model Release: Driving capacity assessment and chaos drills for Context OS and building automated quality gates for model releases. AI Agent Testing: Formulating automated evaluation for AI agents, including multi-turn dialogue and tool calling. Test Engineering & Automation: Developing custom testing tools and integrating automation/performance scripts into CI/CD pipelines. Required qualifications include experience testing large server-side architectures, microservices, and message queues; experience testing LLM, Embedding, RAG, and Agent products; familiarity with vector-retrieval systems (Elasticsearch, Faiss, Milvus) and inference frameworks (vLLM, Triton); proficiency with automation frameworks (Selenium, Playwright, Pytest) and containerization (Kubernetes, Docker); and an established methodology for AI product quality assurance. Benefits include an Employee Stock Ownership Plan (ESOP), access to frontier AI tools, top-spec equipment, and comprehensive medical and insurance coverage.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

Server-side architecture testingMicroservicesMessage queuesLLM testingEmbeddingRAGAI AgentsElasticsearchFaissMilvusvLLMTritonSeleniumPlaywrightPytestKubernetesDockerCI/CDAPI regression testing

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

AI System and Algorithm Evaluation
The role requires designing quality gates for complex AI components, including retrieval and memory systems, embedding drift, and long-tail recall. You must demonstrate a structured methodology for evaluating non-deterministic AI outputs and search relevance.
Server-Side Architecture and Migration QA
You will be responsible for safeguarding core server modules during architecture upgrades. This involves leading API regression testing, ensuring data consistency across microservices, and conducting service-degradation drills.
AI Agent and Model Release Automation
You must build automated quality gates for model releases and formulate evaluation frameworks for AI agents, specifically focusing on multi-turn dialogue, tool-calling accuracy, and error recovery.
Test Engineering and CI/CD Integration
The role demands the development of custom testing tools and the integration of performance and automation scripts into CI/CD pipelines to support rapid, high-velocity releases.

Likely challenges to think through

Evaluating Non-Deterministic AI Behavior
Testing AI agents and RAG systems is inherently different from traditional software testing due to the probabilistic nature of LLMs. Prepare by articulating how you define 'correctness' in multi-turn dialogues and how you measure performance metrics like embedding drift.
Balancing Infrastructure Stability with Rapid Deployment
Plaud is a high-growth company requiring both rigorous stability testing for core server architecture and rapid, automated quality gates for AI model releases. You may face pressure to maintain high quality while supporting a fast-moving CI/CD environment.
Complexity of Vector-Retrieval Systems
The role requires deep familiarity with vector databases like Milvus or Faiss. You may be challenged on your ability to troubleshoot retrieval accuracy and performance bottlenecks within these specialized systems.

Practice your answers

Technical practice

Testing Strategies for RAG Systems

What to coverExplain how you would design quality gates for a RAG-based search system. Focus on metrics for retrieval accuracy, handling embedding drift, and ensuring the relevance of long-tail recall results.

Evaluating AI Agent Reliability

What to coverDescribe your approach to automating the evaluation of AI agents, specifically regarding multi-turn dialogue consistency and tool-calling accuracy. Discuss how you would implement automated error recovery testing.

Performance Testing in Microservices

What to coverHow do you approach performance and stability testing for a microservices architecture that relies on message queues? Mention how you identify bottlenecks and ensure data consistency during high-load scenarios.

Chaos Engineering for AI Infrastructure

What to coverDescribe how you would design and execute chaos drills for a system like Context OS. Explain the importance of these drills in assessing system capacity and resilience during service degradation.

Automated Quality Gates for Model Releases

What to coverDetail the components of an automated quality gate for model releases, including smoke, regression, and canary testing. Explain how you integrate these into a CI/CD pipeline.

Vector Database Testing

What to coverGiven your experience with systems like Milvus, Faiss, or Elasticsearch, how do you validate the integrity and performance of vector-retrieval operations? Discuss the trade-offs between latency and recall accuracy.

API Regression During Architecture Upgrades

What to coverWhen leading an architecture migration, what is your methodology for maintaining a regression matrix? Explain how you ensure data consistency across services during the transition.

Inference Framework Optimization

What to coverHow do you test the performance of inference frameworks like vLLM or Triton? Discuss how you measure throughput and latency to ensure they meet production requirements.

Behavioural practice

Handling Ambiguous AI Quality Requirements

What to coverDescribe a situation where you had to define quality standards for a new or experimental AI feature. How did you determine what 'good' looked like and gain stakeholder alignment?

Managing Technical Debt in Testing

What to coverProvide an example of when you identified a significant gap in testing coverage for a complex system. How did you prioritize addressing this gap while balancing ongoing development velocity?

Cross-Functional Collaboration

What to coverTell me about a time you worked closely with developers to resolve a critical production issue. How did you facilitate the investigation and ensure the fix was robustly tested?

Adapting to Rapid Product Changes

What to coverDescribe a time when a major shift in product direction required you to overhaul your testing strategy. How did you manage the transition and ensure quality was not compromised?

Sharpen your knowledge

Topics to prepare

AI Model and Agent Evaluation Methodology
The role requires designing quality gates for RAG, embedding drift, and multi-turn AI agent interactions. Exercise: Define a testing framework for an AI agent that includes metrics for tool-calling accuracy, response latency, and error recovery in multi-turn dialogues.
Server-Side Architecture and Microservices Testing
The role involves testing large-scale refactors and message queues. Exercise: Create a strategy for performing API regression and data-consistency checks during a hypothetical migration of a core microservice, focusing on maintaining system stability.
AI Infrastructure and Model Release Pipelines
The role requires building automated quality gates for model releases and performing chaos drills. Exercise: Outline a CI/CD pipeline integration that includes canary deployments and automated smoke tests for a new model version.

Skills to revise

Vector Retrieval Systems (Elasticsearch, Faiss, Milvus)
Familiarity with these systems is required for testing recommendation and search quality. Exercise: Practice writing queries or scripts to validate retrieval accuracy and identify potential embedding drift in a sample vector database.
Automation Frameworks (Pytest, Playwright)
Proficiency in these tools is essential for developing custom testing tools and scripts. Exercise: Develop a Pytest suite that automates a functional test case for a REST API, incorporating assertions for response time and data integrity.
Containerization and Orchestration (Docker, Kubernetes)
The role requires testing within containerized environments. Exercise: Write a Dockerfile for a test runner and define a Kubernetes manifest to deploy a performance testing job within a cluster.

Your preparation checklist

Develop an AI Quality Assurance Methodology
Draft a structured approach for evaluating AI-driven products, specifically focusing on how you would measure embedding drift, retrieval accuracy in RAG systems, and the reliability of AI agent tool-calling. Be prepared to explain your strategy for setting up quality gates in a CI/CD pipeline.
Prepare a Technical Case Study on Microservices Testing
Create a brief walkthrough of a past project where you managed testing for a large-scale microservices architecture. Highlight your experience with message queues, API regression testing, and how you ensured data consistency during system migrations or upgrades.
Demonstrate Proficiency with AI Infrastructure Tools
Review your experience with vector databases like Milvus or Faiss and inference frameworks such as vLLM or Triton. Prepare to discuss how you have used these tools to perform capacity assessments or chaos drills in a production-like environment.
Refine Automation Framework Expertise
Be ready to discuss your experience using Pytest, Playwright, or Selenium to automate complex workflows. Specifically, prepare examples of how you have integrated these scripts into CI/CD pipelines to support rapid, reliable model releases.
Formulate Strategic Questions for the Team
Prepare thoughtful questions regarding Plaud's current CI/CD maturity, the specific challenges they face with 'Context OS' testing, and how the QA team collaborates with R&D to define success metrics for new AI features.

Availability last checked: 18 Oct 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in Singapore. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…