Collaborate directly with customers to solve complex challenges
Build autonomous agents using LLMs, planning algorithms, and decision-making frameworks
Implement agent architectures that support autonomy, interactivity, and task completion
Integrate agents into applications, APIs, and workflows (e.g., copilots, chatbots, automation tools)
Connect agents to external services via APIs, databases, and cloud platforms
Tune agent behavior using feedback loops, reinforcement learning, semantic knowledge layer and user interaction
Monitor performance and implement safety, reliability, and guardrail mechanisms
Work cross-functionally with researchers, engineers, and product teams
Maintain clear documentation of agent logic, design decisions, and dependencies
Build and maintain the Enterprise Agents and Tools Registry for metadata and lifecycle management
Implement the Agent Communication Gateway with robust security, rate limits, observability, and cost controls
Ensure agent-level security, including authentication, authorization, and data protection
Optimize cost, scalability, performance, and reliability of agent operations across cloud and on-prem environments
Deploy and customize agentic AI platforms (e.g., LLM agents, orchestration frameworks)
Integrate AI systems with enterprise APIs, data platforms, and workflows
Solve technical blockers across data ingestion, model deployment, and agent behavior
Design and refine prompts to ensure clarity, compliance, and contextual accuracy
Maintain performance metrics and feedback loops for continuous improvement
Build and iterate custom AI solutions tailored to customer needs, leveraging agentic AI frameworks
Own delivery end to end, from scoping to production
Actively contribute to the evolution of Kyndryl’s AI platforms through feedback, code contributions, and collaboration with product teams
Requirements
Bachelor’s degree in computer science, Engineering, or equivalent
Hands-on experience with Python development and frontend UI technologies (e.g., TypeScript, React.js, etc.)
Hands-on experience building AI based solutions using AI frameworks such as LangChain, Microsoft Semantic Kernel, Google ADK or Microsoft Agent Framework
Knowledge of LLMs, AI Agent architectures, Agent Telemetry/Observability frameworks (Langsmith, Langfuse, litellm etc)
Expertise with Docker, Kubernetes, and at least one cloud platform (Azure, AWS, GCP) or on premises
Experience in microservices-based architectures
Solid grasp of the software delivery lifecycle, version control (Git & GitHub), and data engineering tools such as Pandas and Spark
Experience with cloud AI platforms (AWS, Azure, Google AI) and distributed computing architectures
Ability to translate business requirements into technical solutions and communicate technical value to diverse stakeholders, including executive audiences
Hands-on experience with SQL (e.g., PostgreSQL), NoSQL (e.g., MongoDB), and vector databases for agent data storage, semantic queries and retrieval
Proficiency in CI/CD (e.g., GitHub Actions), automated testing, and observability
Proficiency in API development, backend services, and cloud platforms (AWS, Azure, GCP)
Practical knowledge of deploying RAG architectures and integrating structured and unstructured knowledge sources into AI solutions
Familiarity with community-driven AI tools and libraries, including Hugging Face and relevant repositories
Experience designing, building, or integrating multi-agent systems and orchestration frameworks (e.g., LangGraph, Semantic Kernel, Agent Framework, AutoGen, CrewAI)
Knowledge of system-level optimisation and security best practices for scalable AI systems
Willingness to travel up to 25% globally
Tech Stack
AWS
Azure
Cloud
Docker
Google Cloud Platform
JavaScript
Kubernetes
Microservices
MongoDB
NoSQL
Pandas
Postgres
Python
React
Spark
SQL
TypeScript
Benefits
Flexible, supportive environment
Well-being prioritization
Career progression defined from Junior to Principal levels
Access to cutting-edge learning opportunities
Professional development through certifications and coaching