DE&A - AIML - Deep Learning - Generative AI

Job description

**Key Responsibilities** **1. AI Agents for Application Development** - Design and build **AI agents that assist in application development lifecycle (SDLC)** - Develop agents for: - **Code generation & scaffolding** - **API development & integration** - **Code refactoring and optimization** - Enable **developer copilots** for faster feature delivery **2. AI Agents for Application Enhancements** - Build agents to: - Analyze existing codebases and suggest **enhancements or optimizations** - Automate **bug detection and resolution** - Support **impact analysis for changes** - Develop agents for **legacy modernization and code migration (e.g., Java/.NET upgrades)** **3. Testing & QA Automation Agents** - Create agents to: - Automatically generate **unit, integration, and regression test cases** - Perform **test execution and defect prediction** - Enable **self-healing test automation frameworks** **4. LLM & Agent Framework Implementation** - Build solutions using frameworks such as: - **LangChain, LlamaIndex, Semantic Kernel, AutoGen, CrewAI** - Implement: - **Multi-agent orchestration (planner, executor, reviewer agents)** - **Tool-using agents (Git, CI/CD, APIs, databases)** **5. RAG & Context Engineering** - Implement **RAG pipelines** using application code repositories, documentation, and APIs - Build **context-aware agents** using: - Codebases (GitHub, Azure DevOps) - Knowledge repositories (Confluence, SharePoint) **6. DevOps & Integration** - Integrate agents into: - **CI/CD pipelines (Azure DevOps, GitHub Actions)** - **Developer tools (IDE plugins, Copilot extensions)** - Develop APIs/microservices to expose agent capabilities **7. Evaluation & Optimization** - Define metrics for: - Developer productivity improvement - Code quality and defect reduction - Optimize for **cost, latency, and accuracy of LLM usage** **8. Governance & Security** - Ensure: - **Secure code handling and IP protection** - Compliance with **enterprise AI governance** - Guardrails to prevent insecure or non-compliant code generation **Required Skills & Experience** **Core Skills** - Strong programming skills in **Python (mandatory)** and at least one of **Java/.NET/Node.js** - Hands-on experience with **application development & SDLC processes** - Experience with **REST APIs, microservices architecture** **AI / GenAI Skills** - Experience building **AI-powered developer tools or agents** - Strong knowledge of: - **LLMs (OpenAI, Azure OpenAI, open-source models)** - **Prompt engineering & fine-tuning basics** - Experience in **RAG-based solutions** **Agent Frameworks** - Hands-on with: - **LangChain / Semantic Kernel / LlamaIndex** - Exposure to **AutoGen / CrewAI / multi-agent patterns** **DevOps & Tools** - Familiarity with: - **GitHub / Azure DevOps repositories** - **CI/CD pipelines** - **Docker / Kubernetes (preferred)** **Good to Have** - Experience with **GitHub Copilot or similar developer productivity tools** - Exposure to **code analysis tools (SonarQube, SAST/DAST)** - Experience in **legacy modernization projects** - BFSI domain experience (for enterprise use cases) **Experience** - **5–10 years** total experience - **2+ years in GenAI / AI-led development (preferred)**