Gen AI Developer Specialist

HEXAWARE•United States,

About The Position

This role focuses on designing, developing, and deploying Generative AI solutions. The specialist will translate business problems into Gen AI architectures, build prototypes, and evolve them into production services. Key areas include LLM development, retrieval-augmented generation (RAG) pipelines, evaluation and safety measures, MLOps, and ensuring security, privacy, and compliance. Collaboration with various teams is essential for delivering measurable outcomes.

Requirements

Experience with Gen AI architectures (LLM, RAG, agentic patterns, multimodal).
Experience building end-to-end prototypes.
Experience integrating foundation models (hosted APIs or open-source).
Experience with fine-tuning and parameter-efficient methods (e.g., LoRA/QLoRA/PEFT).
Experience with prompt engineering, system messages, tools/functions, and memory strategies.
Experience implementing RAG pipelines (chunking, embeddings, retrieval, re-ranking, filtering).
Experience with vector databases and document stores.
Experience designing data quality checks.
Experience defining evaluation metrics for accuracy, toxicity, bias, and hallucinations.
Experience implementing guardrails and content filters.
Experience with CI/CD, containerization, and IaC.
Experience optimizing inference for latency, throughput, and cost.
Experience with observability tooling.
Experience handling PII securely.
Understanding of data governance, regulatory, and licensing constraints.
Experience collaborating with product, domain SMEs, data engineering, and SRE.
Experience documenting designs, decisions, and runbooks.

Responsibilities

Translate business problems into Gen AI architectures (LLM, RAG, agentic patterns, multimodal).
Build end-to-end prototypes and evolve them into reliable, secure production services.
Select and integrate foundation models (hosted APIs or open-source).
Implement fine-tuning and parameter-efficient methods (e.g., LoRA/QLoRA/PEFT) where needed.
Engineer prompts/system messages, tools/functions, and memory strategies.
Implement RAG pipelines: chunking, embeddings, retrieval, re-ranking, and filtering.
Work with vector databases and document stores; design data quality checks.
Define automated and human-in-the-loop evaluation for accuracy, toxicity, bias, and hallucinations.
Implement guardrails, content filters, and policy enforcement.
Package and deploy services with CI/CD, containerization, and IaC as applicable.
Optimize inference for latency, throughput, and cost; monitor with observability tooling.
Handle PII securely; align with data governance, regulatory, and licensing constraints.
Partner with product, domain SMEs, data engineering, and SRE to deliver measurable outcomes.
Document designs, decisions, and runbooks; share best practices.