AI Engineer - – LLM Orchestration & Voice Agents

Role:

Type:

Experience:

Role Overview

Responsibilities

• Design and implement LLM-based workflows using planner and agent-based patterns
• Build and maintain LangChain / LangGraph pipelines
• Develop and operate voice agents integrating speech-to-text, LLM reasoning, and text-to speech
• Implement memory systems (short-term and long-term) for AI workflows
• Own prompt engineering as versioned, testable artifacts
• Design and run LLM evaluation pipelines (offline + online)
• Add observability: traces, latency, token usage, and failure analysis
• Collaborate closely with backend engineers to deploy AI features to production
• Debug real-world AI failures and continuously improve system quality
  

Required Skills

• Strong experience deploying LLMs in production
• LangChain and/or LangGraph (required)
• Multi-agent systems, planners, routers, and tool calling
• Prompt engineering with structured outputs (JSON / schemas)
• Experience with LLM memory systems and context management
• Retrieval-Augmented Generation (RAG) design and tuning
  

• Designing LLM evaluation frameworks
• Regression testing for AI pipelines
• Automated grading / LLM-as-judge techniques
• Experience with tools like LangSmith or equivalent custom eval systems
  

• Experience building voice agents
• Speech-to-text (e.g., Whisper or equivalent)
• Text-to-speech systems
• Real-time or near–real-time conversational pipelines
• Handling latency, interruptions, and conversational state
  

• Strong Python skills (FastAPI / Flask preferred)
• Async systems, retries, idempotency
• API-first system design
• Working with vector databases and traditional datastores
  

Nice To Have

• Multi-model routing (OpenAI, Anthropic, open-source models)
• Fine-tuning experience (LoRA, adapters, task-specific tuning)
• Cost optimization and token budgeting strategies
• AI observability and monitoring at scale
• Experience shipping AI products to non-technical users