AI Native Engineer – Cognizant
Job Summary:
We are seeking an AI-Native Software Engineer who views AI not just as an autocomplete tool, but as a core collaborative partner in software delivery. In this role, you will spend less time manually writing boilerplate and more time architecting systems, designing precise technical specifications, and orchestrating multi-agent workflows.
Core Responsibilities
- System Architecture & Design: Define high-level system structures, API contracts, and data models before instructing AI tools to implement them. Own the design, not just the execution.
- Context Engineering & Spec Writing: Author rigorous, unambiguous technical specifications and context rules to guide AI agents toward deterministic, reviewable outputs.
- RAG Pipeline Design: Architect and own end-to-end Retrieval-Augmented Generation pipelines, document ingestion, chunking strategy, embedding selection, vector store configuration, hybrid retrieval, and relevance evaluation.
- Agentic Workflow Management: Build and operate agent harnesses using orchestration frameworks (e.g. LangGraph, LangChain, AutoGen) including tool definitions, routing logic, guardrails, fallback paths, and evaluation hooks.
- Human-in-the-Loop Validation: Design and enforce HITL gates for agentic write operations. Know when to automate and when to require human sign-off, especially for irreversible or high-stakes actions.
- Review, test, and audit AI-generated code for security vulnerabilities, performance characteristics, edge cases, and architectural alignment before it reaches production.
Preferred Qualifications
- Bachelor’s or Master’s degree in Computer Science, Software Engineering, or equivalent deep production experience.
- Experience integrating with enterprise HR, workforce, or ERP platforms (e.g. SAP SuccessFactors, Workday, Concur, or Oracle HCM) — particularly in an agentic or API integration context.
- Hands-on ML experience beyond API consumption: model fine-tuning, training pipelines, evaluation frameworks, or MLOps deployment.
- Familiarity with enterprise identity providers (e.g. OKTA, Azure AD) and secure token handling in agentic contexts.
- A portfolio or GitHub repository demonstrating projects built primarily via agentic or spec-driven development methodologies.
Job Details
| Posted Date: | 2026-07-23 |
| Job Location: | United Arab Emirates – Dubai |
| Job Role: | Engineering |
| Company Industry: | Accounting |
