AI/ML Engineer - Multimodal LLM Fine-Tuning and Validation

AI/ML Engineer - Multimodal LLM Fine-Tuning and Validation Company Overview: We are an AI-focused consulting team that designs and delivers production-ready AI systems for clients across a range of industries. We are seeking a hands-on AI/ML Engineer to drive technical execution across multiple active projects. These projects may combine multimodal document processing, structured data extraction, retrieval, business rules, tool use, human-in-the-loop review, and integrations with client platforms. You will work closely with the Head of AI and Head of Product to translate business requirements into reliable, scalable AI solutions. You will own work across the delivery lifecycle, from prototyping and technical design through implementation, evaluation, iteration, and production deployment. We operate in a fast-moving, low-bureaucracy environment where engineers are trusted to take ownership, communicate clearly, and drive work from initial exploration through production delivery. We value practical problem-solving, sound technical judgment, and a strong bias toward action. Core Responsibilities: - Collaborate within a small, agile technical team to ensure seamless integration of AI models into client-facing services and pipelines. - Optimize and validate multimodal large language model (LLM) workflows for structured extraction and prediction tasks. - Evaluate model performance, conducting detailed error analysis and iterative testing to enhance results. - Assist with auditing security concerns in AI solutions and applying guardrails for generating high-fidelity inferences. Required Skills: - Proficiency in Python and SQL. Experience with langchain, langgraph, langfuse, pydantic, and git. - Good understanding of prompt engineering, context engineering, RAG frameworks, chain-of-thought reasoning, ReAct agent patterns. - Good understanding of LLM evaluation pipelines (LLM-as-a-judge) and metrics (factuality & accuracy, relevance, etc.) - Hands-on experience building or optimizing multimodal LLM workflows and agents, particularly for processing unstructured documents, generating qualitative and quantitative inferences, summarization, or sentiment analysis. - Ability to work independently with minimal supervision. Ideal Candidate Traits: - Hands-on and detail-oriented, with the ability to thrive in a fast-moving startup environment that may include some ambiguity. - A “get it done” attitude and proven track record of taking ownership over workstreams. - Comfortable working independently and managing multiple priorities with minimal supervision. Nice-to-Haves: - Experience with document understanding models. - Experience building REST APIs or MCP servers - Familiarity with the AWS or Microsoft Azure ecosystems. - Prior experience working with healthcare, insurance, or financial document data. When submitting your proposal, please include: - A link to your GitHub account or any other portfolio showcasing your work