Yum! Brands’ is hiring AI Engineers to help design and improve voice-based AI agents for Taco Bell drive-thru operations. These roles are perfect for early-career AI engineers or data scientists looking to expand their skills in LLM-based interaction design, speech system optimization, and production-quality prompt development.
You’ll collaborate closely with senior AI Engineers, MLEs, and QA to iterate on agent prompts, tune foundational models, and contribute to the overall agent experience for customers and employees.
Responsibilities
Prompt Engineering & Agent Design
- Author and refine prompt instructions, chaining logic, and fallback strategies
- Design and test multi-turn conversation flows aligned to Taco Bell brand voice
- Build and maintain system personas and error handling routines
Model Tuning & Evaluation
Fine-tune LLMs, ASR models, and embedding systems under supervisionAssist in running experiments using LoRA, distillation, or pruning methodsContribute to agent evaluation metrics, regression tracking, and A / B testsCross-Functional Collaboration
Work closely with MLEs on model integration and performance tuningPartner with QA and PMs to improve agent usability, reliability, and task success ratesHelp manage RAG components, context retrieval chains, and structured data inputsMandatory Skills
4-8 years of experience in AI Engineering, Data Science or ML-related rolesProficiency in Python, SQL and AI frameworks (e.g., LangChain, HuggingFace, OpenAI APIs)Hands-on experience in LLM / NLP fine-tuning, SFT, LoRA, QLoRA, PEFT frameworksStrong experience with RAG developmentAWS proficiency (S3, Lambda, API Gateway, ECS / EKS, possibly SageMaker)Ability to convert models into production-ready applications : - API creation, Microservices, Dockers, CI / CD pipelines, KubernetesExperience in building data / ML pipelines for : transcripts, call logs & conversation data.Comfortable working with US engineering teams (cross-timezone collaboration).Preferred / good to have :
Exposure on ASR / TTS outputs (voice-to-text workflows)Understanding of Conversational AI KPIs (containment, handoff / fallback, AHT impact, etc.)Any experience with real-time orchestration (e.g., routing calls, streaming pipelines)Familiarity with audio / voice analyticsExperience in deploying LLMs in cloud environments beyond AWS (optional).