Job Title: Principal Architect – LLM Agents & Multi-Agent Frameworks
Location: Hyderabad / Chennai
Experience: 14+ years (minimum 5 years in AI/ML)
Job Summary
We are looking for an experienced and forward-thinking Principal Architect to lead the architecture and development of LLM-powered AI Agents and multi-agent frameworks within Cognizant AI Agent Hub. The ideal candidate will combine deep software engineering skills with proven AI/ML experience to deliver high-performing intelligent systems.
Key Responsibilities
Define and evolve architecture for LLM-based intelligent agents and multi-agent systems to ensure scalability, performance, and efficiency.
Design and implement microservices-based architectures that are robust, modular, and easy to maintain.
Integrate LLM agents seamlessly with existing systems, tools, and enterprise data platforms.
Lead development of APIs and SDKs to enable internal and external developers to build upon the platform.
Drive full-stack development using React, Angular, or Vue.js for frontend and Node.js, Python, or Java for backend services.
Architect solutions using cloud platforms such as AWS, Azure, or GCP, with hands-on experience in serverless, containerization, and cloud-native technologies.
Collaborate with data scientists and ML engineers to fine-tune and optimize LLMs for specific business domains and tasks.
Implement model evaluation tools and observability platforms such as LangSmith, Arize, or similar to ensure consistent performance.
Develop prompt engineering strategies to enhance agent output and mitigate hallucination or bias in LLM-generated responses.
Establish ML Ops workflows including CI/CD pipelines, versioning, deployment automation, and experiment tracking.
Design and implement efficient data pipelines to support AI training, inference, and feedback loops, ensuring data governance and security.
Mentor and lead engineering teams, instilling best practices in AI software development and architectural thinking.
Promote a culture of innovation, technical excellence, and continuous improvement across teams.
Stay up to date with the latest developments in LLMs, multi-agent frameworks, and AI tooling, and evaluate their practical applicability to Cognizant’s offerings.
Qualifications
14+ years of software engineering experience with 5+ years in AI/ML system development.
Proven track record in building and deploying LLM-powered applications using tools like GPT, Claude, or similar.
Strong expertise across the full technology stack: frontend, backend, databases, and cloud infrastructure.
Hands-on experience with agent frameworks such as Autogen, LangGraph, AWS Agent Framework, or similar platforms.
Deep understanding of microservices design, data architecture, and enterprise integration strategies.
Experience with MLOps practices, including automated pipelines, model versioning, deployment, and monitoring.
Solid foundation in data engineering including data ingestion, transformation, storage, and governance.
Excellent communication skills with a strong ability to lead discussions with cross-functional teams and stakeholders.
Strong problem-solving mindset, analytical thinking, and a passion for architecting next-generation AI solutions.
Demonstrated ability to lead teams, mentor developers, and scale engineering culture effectively.
About Cognizant
Cognizant is a global professional services company transforming business, operations, and technology models through digital innovation.
Headquartered in the United States, Cognizant serves clients across industries with 300,000+ associates globally.
Cognizant is a member of the NASDAQ-100 and ranked among Forbes World’s Best Employers 2024.
For more information, visit www.cognizant.com
Equal Opportunity Statement
Cognizant is an equal opportunity employer. We consider applicants regardless of race, color, religion, gender, sexual orientation, nationality, disability, veteran status, or other protected status.
Additional Notes
Job descriptions and compensation may change based on business needs and local legal guidelines.
Interviews may be conducted in-person or virtually, and valid ID proof may be required at each stage.
