New opportunity

Performance Engineer

Ascendion · Bengaluru, India

About this role

The Performance Engineer at Ascendion will focus on designing, developing, and testing autonomous AI agents and multi-agent systems to solve complex business problems. This role requires building and optimizing advanced Retrieval-Augmented Generation (RAG) pipelines, managing vector databases, and ensuring enterprise-grade reliability through robust evaluation frameworks and anti-hallucination guardrails. Responsibilities: Design and deploy scalable AI architectures and agentic workflows on Google Cloud Platform (GCP). Integrate LLMs with external APIs and enterprise toolchains. Monitor agent performance, debug reasoning loops, and optimize systems for latency and cost. Collaborate with product managers and data engineers to translate business requirements into technical AI solutions. Requirements: Bachelor's or Master's degree in Computer Science, Artificial Intelligence, Software Engineering, or a related field. Minimum 7 years of experience in Python programming, including asynchronous programming and API development. Proven hands-on experience (6-12 months) building and deploying generative AI applications in production. Deep understanding of LLM capabilities, advanced prompt engineering, and context window management. Strong foundational knowledge of vector search, embeddings, and semantic search. Experience designing complex multi-agent architectures and using orchestration frameworks like LangChain, LlamaIndex, AutoGen, or CrewAI. Background in MLOps, CI/CD for machine learning, and automated model evaluation. Proficiency with containerization (Docker, Kubernetes) and cloud services (GCP Vertex AI, Cloud Run, Cloud Functions, BigQuery) or equivalent AWS/Azure experience. Solid understanding of data security, privacy compliance, and AI governance.

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

Skills for this role

PythonGenerative AILLMPrompt EngineeringVector SearchMulti-agent ArchitecturesLangChainLlamaIndexAutoGenCrewAIMLOpsCI/CDDockerKubernetesGCPVertex AICloud RunCloud FunctionsBigQueryRAG

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

Autonomous Agent Architecture
The role centers on designing and deploying multi-agent systems. You must demonstrate the ability to orchestrate complex workflows where multiple AI models collaborate to solve business problems using frameworks like LangChain, AutoGen, or CrewAI.
Reliability and Anti-Hallucination Guardrails
Enterprise-grade reliability is a core requirement. You must prioritize the implementation of robust evaluation frameworks, grounding techniques, and guardrails to minimize hallucinations in RAG pipelines and agentic outputs.
Production-Grade AI Engineering
Beyond prototyping, you are responsible for the full lifecycle of AI applications. This includes MLOps, CI/CD for machine learning, containerization with Docker/Kubernetes, and optimizing systems for latency and cost on GCP.

Likely challenges to think through

Debugging Non-Deterministic Reasoning Loops
Multi-agent systems often exhibit non-deterministic behavior. The challenge lies in debugging complex reasoning chains. Prepare by discussing your methodology for logging, tracing, and monitoring agent interactions to isolate failures in autonomous workflows.
Balancing Performance with Cost at Scale
Optimizing RAG pipelines and agentic workflows for both latency and cloud costs is a significant technical hurdle. Prepare to explain how you make trade-offs between model complexity, context window management, and infrastructure resource allocation.
Ensuring Governance in Autonomous Systems
Integrating AI with enterprise toolchains requires strict adherence to data security and privacy. You will face the challenge of maintaining compliance while allowing agents to access external APIs and internal databases; be ready to discuss your approach to secure AI governance.

Practice your answers

Technical practice

Designing Multi-Agent Architectures

What to coverExplain your approach to designing multi-agent systems where models collaborate. Discuss how you handle task decomposition, inter-agent communication protocols, and state management to ensure coherent outcomes.

Optimizing RAG Pipelines

What to coverDescribe your methodology for building and optimizing Retrieval-Augmented Generation pipelines. Focus on how you manage vector databases, improve retrieval accuracy, and handle context window limitations.

Implementing Anti-Hallucination Guardrails

What to coverWhat techniques do you use to ensure enterprise-grade reliability and minimize hallucinations in LLM outputs? Discuss grounding techniques, validation frameworks, and how you verify model responses against source data.

Debugging AI Reasoning Loops

What to coverWhen an autonomous agent fails to complete a task or enters an infinite loop, what is your debugging process? Explain how you monitor agent performance and trace reasoning steps to identify bottlenecks or logic errors.

Asynchronous Python for AI Workflows

What to coverGiven the requirement for asynchronous programming in this role, how do you utilize Python's asyncio to handle concurrent API calls and long-running AI tasks? Provide examples of how this improves system latency.

MLOps and CI/CD for GenAI

What to coverHow do you integrate CI/CD pipelines into your generative AI development lifecycle? Discuss your approach to automated model evaluation, version control for prompts, and deployment strategies on platforms like GCP.

Cloud-Native AI Deployment

What to coverDescribe your experience deploying AI architectures on GCP using services like Vertex AI, Cloud Run, or Cloud Functions. Explain how you manage scalability and cost-efficiency in a production environment.

Vector Search and Embeddings

What to coverExplain the trade-offs between different vector search methodologies and embedding models. How do you choose the right indexing strategy to balance search speed and semantic relevance for large-scale datasets?

Behavioural practice

Translating Business Requirements to AI Solutions

What to coverDescribe a time you collaborated with product managers to define a technical AI solution. How did you ensure the final architecture met the business objectives while remaining technically feasible?

Handling AI Governance and Compliance

What to coverProvide an example of how you have addressed data security or privacy compliance in an AI project. How do you balance the need for model performance with strict organizational governance standards?

Managing Technical Trade-offs

What to coverTell me about a situation where you had to choose between system latency, cost, and model accuracy. How did you evaluate the trade-offs and what was the final outcome?

Collaborating in Cross-Functional Teams

What to coverDescribe a challenging project where you worked closely with data engineers or other stakeholders. How did you manage technical disagreements or integration issues to ensure project success?

Sharpen your knowledge

Topics to prepare

Multi-Agent System Architecture
Focus on the design patterns for autonomous AI agents that collaborate to solve complex tasks. Exercise: Draft a high-level architecture diagram for a multi-agent system where one agent handles data retrieval, another processes reasoning, and a third performs validation. This is directly relevant to the requirement for designing complex multi-agent architectures.
Retrieval-Augmented Generation (RAG) Optimization
Deepen understanding of RAG pipelines, specifically focusing on grounding techniques and anti-hallucination guardrails. Exercise: Review strategies for improving retrieval accuracy, such as hybrid search or re-ranking, and document how to implement a guardrail to detect and prevent model hallucinations. This addresses the core responsibility of building and optimizing enterprise-grade RAG pipelines.
AI Governance and Evaluation Frameworks
Study methodologies for automated model evaluation and data privacy compliance in AI. Exercise: Define a set of metrics (e.g., latency, cost, accuracy) to evaluate an agent's performance in a production environment. This is essential for the role's focus on ensuring enterprise-grade reliability and performance monitoring.

Skills to revise

Asynchronous Python Development
Master the use of 'asyncio' and asynchronous API development to handle concurrent AI tasks. Exercise: Refactor a synchronous Python script that calls an LLM API into an asynchronous version to improve throughput. This is a specific requirement for the 7+ years of Python experience needed for high-performance AI systems.
GenAI Orchestration Frameworks
Gain hands-on proficiency with frameworks like LangChain, LlamaIndex, or CrewAI. Exercise: Build a simple agentic workflow using one of these frameworks that integrates an external API and performs a multi-step reasoning task. This directly supports the requirement for practical experience with orchestration tools.
GCP AI Infrastructure Management
Develop skills in deploying and scaling AI applications on Google Cloud Platform using Vertex AI and Cloud Run. Exercise: Create a Dockerfile for a Python-based AI application and outline the steps to deploy it as a containerized service on Google Cloud Run. This aligns with the requirement for cloud-native deployment and containerization.

Your preparation checklist

Prepare a production-grade RAG project walkthrough
Document a project where you built and deployed a RAG pipeline. Focus on how you managed vector databases, handled context window limitations, and implemented anti-hallucination guardrails to ensure reliability.
Demonstrate multi-agent architecture expertise
Prepare to discuss a specific multi-agent system you designed using frameworks like LangChain, AutoGen, or CrewAI. Explain how you managed agent collaboration, reasoning loops, and task delegation.
Review MLOps and evaluation frameworks
Be ready to explain your approach to automated model evaluation and CI/CD for machine learning. Detail how you monitor performance, latency, and cost in production environments.
Align technical skills with GCP services
Map your experience with cloud services to GCP equivalents like Vertex AI, Cloud Run, and BigQuery. Prepare examples of how you have containerized and deployed AI applications using Docker and Kubernetes.
Formulate questions on AI governance and security
Prepare thoughtful questions regarding the company's approach to AI governance, data privacy compliance, and how they balance rapid AI deployment with enterprise-grade security requirements.

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…