NOSHEEN SHEIKH

GenAI, LLM Integration, and Automation

• Architected and implemented LLM-powered backend services using Python and Flask, enabling automation of complex, manual workflows with production-grade reliability.
• Designed prompt engineering strategies, including system prompts, task decomposition, few-shot prompting, chain-of-thought-style reasoning (task-level), prompt templating, and response validation layers to ensure deterministic and high-quality outputs.
• Built LLM orchestration pipelines handling input preprocessing, context enrichment, prompt versioning, output parsing, retries, fallbacks, and guardrails.
• Evaluated and benchmarked Gemini 2.5 Pro Preview, Gemini Flash 2.0, GPT-4o, o1, and o3-mini using latency, cost, accuracy, and consistency metrics; productionized workflows on Gemini 2.5 Pro.
• Implemented AI-driven features, including document classification, semantic summarization, structured content generation, entity extraction, and automated validation.
• Applied Retrieval-Augmented Generation (RAG) concepts using embedding-based semantic search, context chunking, and relevance filtering to improve response accuracy.
• Worked with vector search fundamentals, metadata indexing, and context window optimization for scalable AI workflows.
• Developed agentic workflow patterns involving tool calling, multi-step reasoning, and task delegation across services.
• Prototyped and validated AI workflows using Google AI Studio, OpenAI Playground, and Jupyter Notebook, and then integrated them into production backend systems.

Backend Engineering and System Design

• Designed and maintained Spring Boot-based backend services, focusing on scalability, reliability, and maintainability.
• Developed and exposed RESTful APIs to support file translation, source-to-target data conversion, and backend integrations.
• Built efficient backend systems for large-scale data migration and cloud transformation projects.
• Improved system performance by refactoring legacy codebases, optimizing workflows, and reducing failure points.

Tool Development – Raven (Cloud Migration Support Tool)

• Played a backend-centric role in the development of Raven, focusing on API design, service integration, and query translation workflows.
• Contributed to backend architecture, supporting cloud data warehouse migrations.

Code Reviews and Technical Leadership

• Conducted code and design reviews, ensuring adherence to best practices, performance standards, and maintainability.
• Held merge rights, managing version control, and maintaining code quality across repositories.
• Mentored junior engineers on backend development, system design, and product knowledge.
• Collaborated with cross-functional teams to resolve complex technical challenges, and deliver scalable solutions.