New opportunity

DATA ARCHITECT

TE Connectivity · Bangalore, India

About this role

The Data Architect will lead the end-to-end architecture, design, and deployment of Generative AI, Agentic AI, ML, and deep learning solutions. This role focuses on building scalable data pipelines, retrieval systems, agent orchestration, model serving, and monitoring using platforms like AWS Bedrock, Azure OpenAI, and Databricks. Responsibilities include defining enterprise AI reference architectures, selecting optimal models and patterns, establishing engineering and security standards, and developing multi-agent systems. The architect will also build enterprise-grade RAG solutions, implement prompt engineering and guardrails, and establish LLMOps practices for evaluation and monitoring. Required qualifications include 7+ years of experience in Data Science, Machine Learning, or AI Engineering, with at least 3 years specifically architecting production GenAI/LLM solutions. Candidates must have proven experience with Agentic AI or Multi-Agent systems, hands-on expertise with at least two of the primary cloud platforms (AWS Bedrock, Azure OpenAI, Databricks), and strong proficiency in Python, SQL, and PySpark. Experience deploying AI solutions via FastAPI, Docker, and Kubernetes is required. Preferred qualifications include knowledge of LLM fine-tuning (LoRA/PEFT), evaluation frameworks, and exposure to domains like Knowledge Graphs, Document AI, or financial risk modeling. TE Connectivity offers a competitive salary package, performance-based bonuses, health and wellness incentives, and an employee stock purchase program.

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

Skills for this role

Generative AIAgentic AIMachine LearningDeep LearningAWS BedrockAzure OpenAIDatabricksLangGraphLangChainCrewAIAutoGenSemantic KernelRAGPrompt EngineeringLLMOpsPythonSQLPySparkFastAPIDockerKubernetesCI/CDMLflow

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

Enterprise AI Reference Architecture
The role requires defining and owning the end-to-end architecture for Generative and Agentic AI. You must demonstrate the ability to design scalable systems that integrate data ingestion, retrieval, model serving, and security across AWS Bedrock, Azure OpenAI, and Databricks while balancing latency, cost, and compliance.
Agentic AI and Multi-Agent Orchestration
A core priority is the design and implementation of multi-agent systems using frameworks like LangGraph, CrewAI, or AutoGen. You must be prepared to discuss how to manage agent communication, tool calling, and state management in complex, production-grade workflows.
LLMOps and Evaluation Frameworks
You are expected to establish robust LLMOps practices, including groundedness validation, prompt versioning, and tracing. This involves building systems that not only deploy models but also continuously monitor performance, cost, and output quality in a production environment.
Scalable Data and RAG Engineering
You must build enterprise-grade Retrieval-Augmented Generation (RAG) solutions. This requires expertise in advanced chunking, hybrid search, semantic search, and re-ranking, supported by scalable data pipelines built with PySpark and Databricks.

Likely challenges to think through

Balancing Model Performance and Cost
The role requires selecting optimal models across multiple cloud platforms. The challenge lies in justifying architectural choices that optimize for accuracy and latency without incurring excessive costs. Prepare by being ready to discuss trade-offs between proprietary LLMs and smaller, cost-effective models for specific use cases.
Managing Multi-Agent Complexity
Designing multi-agent systems introduces risks regarding non-deterministic behavior and debugging difficulty. Prepare by articulating strategies for implementing guardrails, structured outputs, and robust logging to maintain system reliability when agents interact autonomously.
Standardizing AI Engineering Practices
As a lead architect, you must establish engineering and security standards across diverse teams. The challenge is ensuring consistent coding, security, and deployment practices (via Docker/Kubernetes) while maintaining the flexibility needed for rapid AI experimentation.

Practice your answers

Technical practice

Designing Multi-Agent Systems

What to coverExplain your approach to orchestrating multi-agent systems using frameworks like LangGraph or AutoGen. Focus on how you manage agent communication, task decomposition, and state management to ensure reliability in complex workflows.

Enterprise RAG Architecture

What to coverDescribe how you design scalable RAG pipelines, specifically addressing chunking strategies, hybrid search, and re-ranking. Discuss how you ensure data retrieval accuracy and handle context window limitations in production environments.

LLMOps and Evaluation Frameworks

What to coverDetail your strategy for implementing LLMOps, including prompt versioning, groundedness validation, and automated evaluation. Suggestion: Prepare to discuss how you measure model performance beyond simple accuracy, such as latency, cost, and hallucination rates.

Cloud Platform Trade-offs

What to coverCompare the architectural patterns and service offerings of AWS Bedrock versus Azure OpenAI. Explain how you select the optimal platform based on specific enterprise requirements like data residency, model availability, and integration with existing infrastructure.

Scalable Data Pipelines with PySpark

What to coverDiscuss your experience building feature engineering pipelines using PySpark and Databricks. Focus on how you optimize performance for large-scale datasets and ensure data quality within a Delta Lake architecture.

Model Serving and Deployment

What to coverExplain your process for deploying AI models as scalable APIs using FastAPI, Docker, and Kubernetes. Address how you handle container orchestration, load balancing, and CI/CD integration for production-grade AI services.

Implementing Guardrails and Security

What to coverDescribe how you implement security standards and guardrails for LLM applications. Discuss techniques for preventing prompt injection, managing PII, and ensuring structured outputs for downstream system integration.

Fine-tuning vs. In-context Learning

What to coverExplain the trade-offs between using RAG, prompt engineering, and fine-tuning (LoRA/PEFT) for domain-specific tasks. Suggestion: Be prepared to justify when fine-tuning is necessary versus when prompt-based approaches are sufficient for business needs.

Behavioural practice

Leading Technical Strategy

What to coverDescribe a situation where you defined an enterprise-level architecture for a complex AI project. Focus on how you balanced competing priorities like cost, latency, and scalability while gaining stakeholder buy-in.

Handling Technical Debt in AI

What to coverProvide an example of how you identified and addressed technical debt in a production machine learning or AI system. Explain the steps you took to refactor the solution without disrupting ongoing business operations.

Cross-functional Collaboration

What to coverDescribe a time you had to explain complex AI architectural decisions to non-technical stakeholders. How did you ensure they understood the risks and benefits associated with your proposed solution?

Managing Project Failure

What to coverShare an experience where an AI deployment or model performance did not meet expectations. Detail how you diagnosed the root cause, communicated the issue, and implemented a corrective action plan.

Sharpen your knowledge

Topics to prepare

Agentic AI Orchestration
Review the design patterns for multi-agent systems, specifically focusing on how to manage state, tool-calling, and inter-agent communication. Exercise: Draft a high-level architecture diagram for a multi-agent system that handles complex document retrieval and decision-making, ensuring you can explain the choice of framework (e.g., LangGraph vs. AutoGen) based on scalability and latency requirements.
Enterprise LLMOps and Evaluation
Study the lifecycle of production LLM applications, focusing on evaluation frameworks, groundedness validation, and monitoring. Exercise: Define a set of KPIs and a testing strategy for a RAG pipeline, including how you would measure retrieval accuracy, hallucination rates, and cost-efficiency in a production environment.
Enterprise AI Reference Architecture
Understand the integration of AI solutions within existing enterprise data ecosystems. Exercise: Create a technical decision record (TDR) comparing the trade-offs between using AWS Bedrock and Azure OpenAI for a hypothetical enterprise use case, considering security, compliance, and existing data infrastructure.

Skills to revise

RAG Pipeline Development
Practice building advanced RAG systems using vector databases and hybrid search. Exercise: Implement a pipeline that incorporates chunking strategies, semantic search, and re-ranking, and demonstrate how you would integrate this with a vector database like Databricks or similar.
Scalable Data Engineering with PySpark
Focus on building robust data pipelines for AI/ML workloads. Exercise: Write a PySpark script to process and clean a large dataset, ensuring it is optimized for feature engineering and ready for consumption by an LLM or ML model.
Containerized Deployment (Docker/Kubernetes)
Ensure proficiency in packaging and deploying AI services. Exercise: Create a Dockerfile for a FastAPI-based AI service and write a basic Kubernetes deployment manifest, including health checks and resource limits suitable for a production environment.

Your preparation checklist

Prepare an Agentic AI Architecture Case Study
Develop a technical presentation or document detailing a production-grade multi-agent system you designed. Focus on the orchestration framework used (e.g., LangGraph or AutoGen), how you handled tool calling, and the specific strategies implemented for agent communication and state management.
Map Your Experience to Cloud-Native AI Platforms
Create a comparison table highlighting your hands-on experience with AWS Bedrock, Azure OpenAI, and Databricks. Prepare to discuss how you select between these platforms based on specific enterprise requirements like latency, cost, and security guardrails.
Document Your RAG and LLMOps Pipeline
Prepare a walkthrough of an end-to-end RAG solution you built. Include details on your chunking strategy, vector database selection, and the specific LLMOps practices you established for evaluation, tracing, and monitoring to ensure groundedness and performance.
Formulate Strategic Questions for the Interviewer
Prepare questions regarding TE Connectivity's current AI maturity, such as how they balance model selection between proprietary and open-source, their approach to data governance in multi-agent environments, and how they integrate AI solutions into existing engineering data domains like PDM or SAP.

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 India. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…