New opportunity

Agentic AI Intern (Hybrid) - Able to Commit 5 - 6 Months

TE Connectivity · Singapore

About this role

Role Purpose: TE Connectivity is seeking a technically strong and motivated Agentic AI Intern to join the Corporate R&D Center in Singapore. You will be embedded in a newly formed AI team building TE’s AI Hub, working on real-world problems at the intersection of frontier AI and industrial engineering. This is an active, hands-on role where you will design and build agentic AI systems and contribute to a platform used by global business units. Responsibilities: Design, implement, and evaluate agentic AI workflows using orchestration frameworks like LangChain, LlamaIndex, AutoGen, or CrewAI. Build multi-agent systems incorporating tool-use, RAG, memory management, and multi-step planning. Develop RAG pipelines with vector databases, benchmark LLMs (GPT-4o, Claude, Gemini, etc.), and apply prompt engineering techniques. Write clean, testable Python code, manage version control with Git, and assist with cloud deployment (AWS, Azure, GCP) and containerization. Collaborate with R&D scientists and local research partners to embed AI into design and manufacturing workflows. Requirements: Currently pursuing a Bachelor’s or Master’s degree in Computer Science, AI, Data Science, or related fields. Solid Python programming skills and a strong understanding of LLM capabilities and agentic concepts. Hands-on experience with at least one agentic framework. Strong analytical mindset and ability to work with ambiguous problems. Preferred Qualifications: Experience with RAG systems, vector databases, model evaluation frameworks (RAGAS, LLM-as-a-judge), and cloud platforms. Knowledge of responsible AI practices, ML libraries (PyTorch, Hugging Face), and interest in industrial engineering domains. Benefits: Hands-on experience in a corporate R&D environment, mentorship from AI scientists, and opportunities to collaborate with Singaporean research institutes.

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

Skills for this role

PythonLangChainLlamaIndexAutoGenCrewAIRAGFAISSChromaWeaviateLLMsPrompt EngineeringGitAWSAzureGCPDockerPyTorchHugging Face Transformers

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

Design and Implementation of Agentic Workflows
The primary focus is building multi-agent systems using frameworks like LangChain, LlamaIndex, or AutoGen. You must demonstrate the ability to integrate tool-use, memory management, and multi-step planning to solve specific industrial engineering problems.
Development of Robust RAG Pipelines
You are expected to build and integrate retrieval-augmented generation pipelines with vector databases to enable semantic search over technical engineering documentation, ensuring high accuracy and reliability in knowledge retrieval.
Systematic Benchmarking and Evaluation
You must apply structured evaluation methods to assess LLM performance for domain-specific tasks. This includes using frameworks like RAGAS or LLM-as-a-judge to validate outputs, mitigate hallucinations, and ensure safe deployment.
Software Engineering and Cloud Integration
You are required to write clean, testable Python code and manage version control via Git. You will also assist in deploying these AI agents into cloud environments (AWS, Azure, or GCP) using containerization tools.

Likely challenges to think through

Bridging Frontier AI with Industrial Engineering
A significant challenge is applying abstract AI capabilities to concrete, complex manufacturing and product design workflows. Prepare by researching how AI agents can handle domain-specific constraints and technical data, rather than just general-purpose tasks.
Managing Ambiguity in R&D Environments
As this is an R&D role, you will likely face ill-defined problems without clear solutions. Prepare by demonstrating a structured approach to experimental design, where you can break down ambiguous requirements into testable hypotheses and iterative development cycles.
Ensuring Reliability and Safety in AI Outputs
Industrial applications require high precision. You will face the difficulty of balancing agent autonomy with the need for guardrails. Prepare by discussing how you implement validation layers and responsible AI practices to prevent errors in critical engineering contexts.

Practice your answers

Technical practice

Designing Multi-Agent Workflows

What to coverExplain how you would structure a multi-agent system using frameworks like AutoGen or CrewAI to solve a complex engineering task. Discuss how you would manage task decomposition, inter-agent communication, and the delegation of specific tools to different agents.

RAG Pipeline Optimization

What to coverDescribe the components of a robust RAG pipeline for technical engineering documentation. Address how you would handle document chunking strategies, vector database selection (e.g., FAISS or Chroma), and retrieval accuracy improvements.

LLM Benchmarking and Evaluation

What to coverHow would you design an evaluation framework to assess the reliability and accuracy of an AI agent in an industrial context? Discuss the use of metrics or techniques like RAGAS or LLM-as-a-judge to validate agent outputs against ground truth.

Prompt Engineering for Complex Tasks

What to coverCompare few-shot prompting, chain-of-thought, and structured output generation in the context of agentic workflows. Explain when you would prioritize one technique over another to improve agent performance and reliability.

Memory Management in Agents

What to coverDiscuss the challenges of maintaining state and context in long-running agentic workflows. Explain how you would implement memory management to ensure an agent retains relevant information across multiple steps without exceeding token limits.

Cloud Deployment and Containerization

What to coverDescribe your process for containerizing an AI application using Docker for deployment on cloud platforms like AWS or Azure. Mention how you would handle environment variables, dependency management, and basic CI/CD considerations.

Responsible AI and Guardrails

What to coverWhat strategies would you implement to mitigate hallucinations and ensure safe deployment of AI agents in a manufacturing environment? Discuss the importance of output validation and implementing guardrails to prevent unintended agent behavior.

Python Software Engineering Best Practices

What to coverHow do you ensure your code is testable and maintainable when working in a collaborative R&D environment? Address your approach to modular design, documentation, and using Git for version control in a team setting.

Behavioural practice

Navigating Ambiguity in R&D

What to coverDescribe a time when you were faced with a technical problem that lacked clear instructions or a defined solution. Explain the steps you took to research, experiment, and eventually arrive at a viable approach.

Collaborating Across Disciplines

What to coverProvide an example of a project where you had to explain a complex technical concept to a non-technical stakeholder or collaborator. How did you adapt your communication style to ensure alignment and project success?

Handling Failure in Experiments

What to coverTell me about a time when an experiment or a technical implementation did not yield the expected results. How did you analyze the failure, what did you learn, and how did you pivot your strategy?

Managing Multiple Priorities

What to coverDescribe a situation where you had to balance multiple tasks or project requirements simultaneously. How did you prioritize your work to ensure you met deadlines while maintaining code quality?

Sharpen your knowledge

Topics to prepare

Multi-Agent Orchestration and Planning
Understand how to design systems where multiple agents collaborate using tool-use, memory, and multi-step planning. Exercise: Build a small-scale system using AutoGen or CrewAI where one agent retrieves technical data and another summarizes it for an engineering report, focusing on how they hand off tasks.
Retrieval-Augmented Generation (RAG) Pipelines
Master the integration of vector databases with LLMs to ground outputs in domain-specific technical documentation. Exercise: Create a pipeline that indexes a set of technical PDFs into a vector database like FAISS or Chroma and implement a query interface that cites sources in its response.
LLM Evaluation and Benchmarking
Learn to assess model performance using structured frameworks to ensure reliability in industrial contexts. Exercise: Use RAGAS or a custom 'LLM-as-a-judge' script to evaluate the accuracy and hallucination rate of a RAG pipeline against a set of ground-truth engineering questions.

Skills to revise

Python for Agentic Frameworks
Develop clean, modular Python code using LangChain or LlamaIndex. Exercise: Refactor a simple script into a modular class structure that utilizes LangChain Expression Language (LCEL) to manage agent state and tool execution.
Prompt Engineering for Structured Output
Apply techniques like few-shot prompting and chain-of-thought to ensure agents produce reliable, machine-readable data. Exercise: Design a prompt template that forces an LLM to output JSON-formatted technical specifications, including validation logic to handle parsing errors.
Containerization and Deployment Basics
Prepare AI applications for cloud environments using Docker. Exercise: Create a Dockerfile for a Python-based agent application, ensuring all dependencies are pinned and the environment is reproducible for deployment to a cloud provider.

Your preparation checklist

Build a Multi-Agent Prototype
Develop a project using LangChain, AutoGen, or CrewAI that demonstrates multi-step planning and tool-use. Ensure the code is modular, hosted on GitHub, and includes a README explaining how the agents coordinate to solve a specific task.
Construct a RAG Pipeline with Evaluation
Create a RAG system using a vector database like FAISS or Chroma. Incorporate an evaluation framework such as RAGAS or LLM-as-a-judge to measure retrieval accuracy and hallucination rates, documenting your benchmarking process.
Prepare a Technical Walkthrough
Select one of your AI projects and prepare a 5-minute explanation focusing on your design choices, how you handled ambiguity, and the specific prompt engineering techniques (e.g., chain-of-thought) used to improve performance.
Review Software Engineering Fundamentals
Be ready to discuss your experience with Git version control, writing testable Python code, and containerization with Docker. Prepare examples of how you have debugged complex workflows or managed dependencies in previous projects.
Formulate Targeted Questions for the Team
Prepare questions regarding the R&D team's current approach to model evaluation, how they handle data privacy in industrial engineering contexts, and the specific cloud infrastructure (AWS, Azure, or GCP) they prioritize for agent deployment.

Availability has not yet been verified. 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…