AI Engineer

nttdata

Bengalor NM Years Exp Posted 1h ago

Design and develop AI and Generative AI applications using Large Language Models (LLMs) and modern AI frameworks.
Build and integrate AI agents and agentic workflows capable of reasoning, tool calling, task execution, and interaction with enterprise systems.
Develop prompt engineering, context management, structured-output, and function/tool-calling capabilities.
Implement Retrieval-Augmented Generation (RAG) solutions using enterprise data sources, embeddings, semantic search, and vector databases.
Develop reusable AI components, APIs, services, connectors, and accelerators.
Integrate AI capabilities into web applications, enterprise platforms, workflows, and existing business systems.

Integrate commercial and open-source LLMs using APIs and model-serving platforms.
Develop single-agent and multi-agent workflows using appropriate orchestration frameworks.
Implement agent tools, memory, state management, workflow orchestration, and human-in-the-loop capabilities.
Implement mechanisms to improve grounding and reduce hallucinations and unreliable responses.
Develop AI guardrails, input/output validation, content controls, and error-handling mechanisms.
Optimize prompts, model parameters, retrieval strategies, and workflows for quality, latency, reliability, and cost.

Build ingestion and indexing pipelines for structured and unstructured enterprise content.
Implement document parsing, chunking, metadata enrichment, embedding generation, indexing, and retrieval.
Develop semantic, keyword, and hybrid search capabilities.
Integrate vector databases and enterprise search platforms.
Optimize retrieval quality through query transformation, reranking, filtering, and contextual retrieval techniques.
Implement source attribution and grounding mechanisms where required.

Develop REST APIs, microservices, event-driven integrations, and backend services supporting AI applications.
Integrate AI solutions with enterprise applications, databases, APIs, SaaS platforms, and external services.
Implement authentication, authorization, secrets management, and secure API communication.
Develop connectors and tools that enable AI agents to securely interact with enterprise systems.

Develop automated tests for AI applications, APIs, workflows, and integrations.
Build evaluation datasets and test scenarios for LLM and RAG solutions.
Evaluate AI outputs for accuracy, relevance, groundedness, completeness, safety, and consistency.
Perform prompt, model, retrieval, and agent workflow evaluations.
Troubleshoot AI application issues and perform root-cause analysis.
Support performance, scalability, resilience, and security testing.

Deploy AI applications and services into development, test, staging, and production environments.
Implement CI/CD pipelines for AI applications and supporting services.
Containerize applications using Docker and deploy to Kubernetes, serverless, or cloud-native environments.
Implement model, prompt, configuration, and application version management.
Support LLMOps/MLOps processes including deployment, monitoring, evaluation, and lifecycle management.
Implement logging, tracing, metrics, token/cost monitoring, and AI observability.

Implement security controls defined by solution architecture and organizational standards.
Protect sensitive and personally identifiable information used by AI applications.
Implement safeguards against prompt injection, data leakage, unauthorized tool execution, and inappropriate AI outputs.
Support Responsible AI requirements including transparency, traceability, auditability, and human oversight.

Work closely with AI Technical Architects to implement approved solution architectures and design patterns.
Participate in technical design sessions, code reviews, architecture reviews, and troubleshooting sessions.
Collaborate with data, application, cloud, cybersecurity, and infrastructure teams.
Produce technical documentation covering application design, APIs, integrations, deployment, configuration, and operational procedures.
Identify technical risks, dependencies, and implementation constraints and escalate them appropriately.<