Senior Artificial Intelligence Engineer

Position Title:

About the Role

We are seeking exceptional Senior AI Engineers to lead the development and optimization of agentic AI systems powering Velogent-Aivar's flagship agentic process automation platform for regulated industries. You will architect advanced multi-agent systems that reason, validate, and act autonomously on complex business processes like invoice processing, RFQ management, and loan approvals while maintaining strict regulatory compliance. Your work will directly impact enterprise customers who require 80%+ reduction in manual effort with 90%+ accuracy.

Core Responsibilities

• Design and architect advanced multi-agent systems using LLM reasoning frameworks capable of autonomously executing complex, multi-step business processes
• Develop sophisticated prompt engineering strategies and agent orchestration patterns enabling agentic AI to reason through domain-specific business logic with explainability
• Implement various multi-agentic patterns including critic agents, observer agents, and specialized task agents within the ReVAct (Reason, Validate, Act) framework
• Build and optimize LLM reasoning pipelines supporting chain-of-thought, tree-of-thought, and graph-based reasoning approaches for complex decision-making
• Design validation frameworks ensuring agentic AI decisions maintain regulatory compliance and audit trail requirements in financial services, healthcare, and life sciences
• Implement guardrails and inspection systems detecting emergent agent behaviors and enforcing business rule boundaries
• Develop evaluation frameworks and benchmarking methodologies for assessing multi-agent system accuracy, consistency, and decision quality
• Lead integration with enterprise systems including document repositories, APIs, email systems, and knowledge bases through Model Context Protocol (MCP) standards
• Mentor junior engineers and establish best practices for agentic AI development and governance
• Stay current with latest LLM advances (GPT-4, Claude, Llama) and agentic frameworks (LangChain, CrewAI, Autogen, etc.)

Must-Have Qualifications

• LLM Reasoning Expertise:

  Deep knowledge of large language model capabilities, limitations, and advanced reasoning techniques (chain-of-thought, few-shot prompting, in-context learning)

• Multi-Agent Architecture:

  Production experience designing and building multi-agent systems with understanding of agent coordination, communication patterns, and task decomposition

• Agentic AI Frameworks:

  Hands-on experience with agentic AI implementations using frameworks like LangChain, CrewAI, AutoGen, or similar; ability to architect complex agent interactions

• LLM Fine-Tuning & Optimization:

  Experience fine-tuning LLMs for domain-specific tasks, adapting models for regulatory compliance, and optimizing for inference efficiency

• LangFuse / Observability:

  Proficiency with LLM observability platforms (LangFuse, LangSmith, or equivalent) for debugging, monitoring, and optimizing multi-agent workflows

• Python Expertise:

  Advanced Python skills for implementing complex AI systems, agent frameworks, and data processing pipelines

• Production AI Systems:

  Experience deploying AI/ML systems at scale with focus on reliability, monitoring, and operational governance

• LLM API Integration:

  Deep experience working with leading LLM APIs (OpenAI, Anthropic, AWS Bedrock, open-source models) in production environments

Nice-to-Have Qualifications

• Experience with document processing pipelines and information extraction from unstructured data (PDFs, contracts, financial documents)
• Knowledge of reinforcement learning from human feedback (RLHF) or similar techniques for refining agent behavior
• Background in natural language processing, computational linguistics, or formal reasoning systems
• Familiarity with regulated industry workflows (financial services, healthcare, life sciences) and compliance requirements
• Experience building knowledge graphs, semantic reasoning systems, or ontology-based AI applications
• Understanding of uncertainty quantification and confidence scoring in AI predictions
• Prior experience with rule-based systems or expert systems that could inform agentic design
• Published research or open-source contributions in LLM reasoning or multi-agent systems
• Knowledge of interpretable AI and explainability techniques for regulated environments

What You'll Work With

• Advanced LLM platforms and APIs (GPT-4, Claude, Llama, Bedrock)
• Agentic AI frameworks (LangChain, CrewAI, AutoGen, Haystack)
• Document processing and knowledge management systems
• LLM observability tools (LangFuse, LangSmith)
• Multi-tenant SaaS backend infrastructure on AWS
• Enterprise integration patterns (APIs, email, document repositories)
• Model Context Protocol (MCP) for system integration

Application & Next Steps

Qualified candidates will participate in:

1. Initial screening conversation with hiring manager
2. Technical assessment or portfolio review aligned with role-specific focus
3. Collaborative technical interview with team members
4. Final conversation with engineering leadership