AI LLM Architecture Senior Analyst

h3About The Role /h3 pAs a hands-on AI Engineer, you will be at the heart of designing and developing the building blocks that power advanced AI systems for modern enterprises. This is a highly technical, implementation-focused engineering role where you will spend most of your time designing, developing, integrating, and testing AI system components across traditional machine learning, generative AI, and agentic systems as part of active client engagements. /p pYou will transform detailed architectures and design specifications into production-ready software components. This includes writing clean, maintainable code, making low-level design decisions within your area of responsibility, and ensuring your components integrate reliably into broader AI systems. /p pYou will build and assemble various AI agent system components, including: /p ul liIndividual agent logic /li liTool integrations /li liSkills and capability components /li liMemory components /li /ul pYou will also contribute to the development and integration of foundation models and traditional machine learning models into end-to-end AI pipelines. /p pA practical curiosity for the open-source ecosystem is essential. You will continuously evaluate, learn, and adopt relevant libraries and frameworks covering areas such as: /p ul liAgent orchestration /li liVector storage /li liModel serving /li liMachine Learning pipelines /li /ul pYou will select and apply the most appropriate technologies for each business challenge. Additionally, you will configure, integrate, and deploy third-party AI technologies and platform services, developing a solid understanding of their capabilities and limitations to effectively leverage them within large-scale enterprise environments. /p pYou will design enterprise-grade components that meet requirements for: /p ul liSecurity /li liObservability /li liGovernance /li liPerformance /li liScalability /li /ul pYou will create and maintain technical deliverables related to your engineering work, including: /p ul liLow-level design documents /li liComponent specifications /li liIntegration contracts /li /ul pThis ensures that your developments are well documented, testable, and easily transferable. /p pYou will work as a technical contributor within multidisciplinary teams comprising Data Engineers, ML Engineers, and Application Developers. You will operate under the guidance of Lead Architects and Principal Architects while actively contributing to technical problem-solving and design discussions within your area of expertise. /p pThis role represents an excellent opportunity to develop deep hands-on expertise across the AI Engineering stack, strengthen your software engineering capabilities for cutting-edge AI systems, and progressively grow into a Lead Engineer or Architect position. /p h3Key Responsibilities /h3 h3AI Agents and Multi-Agent Systems /h3 ul liDesign, develop, and configure individual agents, including their prompts, tools, and skills, and integrate them into multi-agent workflows. /li liImplement agent orchestration logic to manage task handoffs, communication, and error recovery. /li liDevelop prompt assembly logic and manage the information provided to models within their context windows. /li liImplement memory components that store and retrieve conversational history and relevant context. /li /ul h3Foundation Models and Fine-Tuning /h3 ul liIntegrate foundation models into applications by selecting the most appropriate model and invocation strategy for each use case. /li liDesign and execute fine-tuning pipelines, including data preparation and model training, to adapt models for specific business domains. /li liApply a strong understanding of Transformer-based architectures. /li /ul h3Retrieval-Augmented Generation (RAG) /h3 ul liDevelop ingestion pipelines capable of parsing, chunking, enriching, and indexing unstructured enterprise content for search and retrieval purposes. /li liImplement embedding generation, integrate vector databases, and develop retrieval components, including connectors and adapters required to support end-to-end RAG (Retrieval-Augmented Generation) pipelines. /li /ul h3Evaluation and Quality Assurance /h3 ul liDesign evaluation frameworks and test suites to measure agent and component quality using metrics such as: Accuracy Relevance Faithfulness /li liShare findings and recommendations to drive continuous design improvements. /li /ul h3Security, Governance, and Compliance /h3 ul liImplement guardrails, content filtering mechanisms, and protections against prompt injection attacks. /li liDevelop PII (Personally Identifiable Information) detection and redaction components. /li liIntegrate access-control mechanisms governing model and tool usage. /li liImplement versioning, audit logging, and lineage tracking while maintaining documentation that supports full system auditability. /li /ul h3Observability and Operations /h3 ul liInstrument components with logging and trac