AI LLM Architect

<b>Description</b><br><p>Fetcherr experts in deep learning, e-commerce, and digitization, Fetcherr disrupts traditional systems with its cutting-edge AI technology. At its core is the Large Market Model (LMM), an adaptable AI engine that forecasts demand and market trends with precision, empowering real-time decision-making. Specializing initially in the airline industry, Fetcherr aims to revolutionize industries with dynamic AI-driven solutions.</p><p>We are seeking an experienced <strong>LLM (AI) Architect</strong> to lead the design and implementation of a <strong>production-grade, LLM-powered question-answering and graph plotting system</strong> that allows users to interact with complex internal data using natural language.</p><p>You will not train foundation models—but instead <strong>orchestrate LLM-powered architectures</strong> (e.g. using OpenAI, Claude, Gemini, etc.), focused on <strong>retrieval-augmented generation (RAG)</strong>, <strong>prompt engineering</strong>, and <strong>context-aware querying</strong> across structured and unstructured internal data sources.</p><p><strong>Responsibilities</strong></p><p><strong>LLM-Powered System Design</strong></p><ul><li>Design and build systems that let users <strong>query internal data</strong> using natural language, simulating an <strong>AI analyst</strong>.</li><li>Create robust pipelines that use <strong>LLMs + internal structured/unstructured data</strong> to provide accurate and explainable responses.</li><li>Architect and optimize <strong>RAG systems</strong>&nbsp;</li></ul><p><strong>Prompt Engineering &amp; Tooling</strong></p><ul><li>Develop advanced prompt strategies for dynamic querying, chaining, and task delegation.</li><li>Implement fallback strategies, guardrails, and context control for reliability and consistency.</li><li>Tune prompts and system behavior to balance <strong>accuracy, latency, and cost</strong>.</li></ul><p><strong>Infrastructure &amp; Deployment</strong></p><ul><li>Work with data engineers and MLOps to <strong>deploy and scale LLM-based services</strong> in production.</li><li>Integrate <strong>vector databases</strong>, embedding pipelines, and caching layers to optimize performance.</li><li>Ensure systems are <strong>monitored, observable, and cost-aware</strong>.</li></ul><p><strong>Collaboration &amp; Productization</strong></p><ul><li>Partner with product managers and analysts to define <strong>use cases</strong> and measure business impact.</li><li>Translate user needs and business logic into scalable LLM-powered applications.</li><li>Educate internal teams on the capabilities and limitations of LLMs in the company context.</li></ul><br> <b>Requirements</b><br><p><strong>You'll be a great fit if you have…</strong></p><ul><li>5+ years of experience in machine learning, AI engineering, or backend systems.</li><li>2+ years working specifically with <strong>LLM architectures or generative AI applications</strong>.</li><li>Hands-on experience with:</li><li><strong>RAG frameworks</strong> (LangChain, LlamaIndex, etc.)</li><li><strong>Embedding models</strong> and pipelines</li><li>LLM APIs (OpenAI, Claude, Gemini, etc.)</li><li>Strong Python skills and familiarity with cloud infrastructure (GCP preferred).</li><li>Proven track record building <strong>reliable, production-grade AI systems</strong>.</li><li>Fluent in <strong>English</strong> (spoken and written) for documentation and cross-team collaboration.</li></ul><p><strong>Mindset &amp; Approach</strong></p><ul><li>Deeply product-oriented with a strong user empathy.</li><li>Balances experimentation with engineering discipline.</li><li>Collaborative, hands-on, and outcome-driven.</li></ul><p><strong>Nice to Have:</strong></p><ul><li>Experience in <strong>analytics, BI, or data exploration</strong> interfaces.</li><li>Familiarity with <strong>semantic search, question decomposition</strong>, and tool-augmented LLMs.</li><li>Background in <strong>pricing, forecasting</strong>, or <strong>airline data</strong> domains.</li></ul><br>