AI Architect

We are looking for an experienced AI Architect to join our progressive team.You’ll work with a multi-service platform designed for financial analysis, portfolio construction, and AI-assisted investment advisory. The platform combines a Snowflake-backed data lake with a Neo4j dual-graph architecture (domain ontology and lexical/GraphRAG), enabling unified querying of portfolios, financial instruments, and unstructured research through a single financial domain-specific language (DSL). The system supports both streaming chat and voice interfaces.ResponsibilitiesOwn and drive the end-to-end architecture, from raw data ingestion to AI-powered advisory outputsDefine and enforce clear contracts across all layers, including data assets, ontology and graph structures, financial DSL, LLM gateway, tool gateway, agent layer, and user experienceDesign systems with strong governance, ensuring all components are backed by documented architectural decisions (ADR/RFC) and traceable justificationBuild and deliver hands-on solutions, iterating incrementally while maintaining architectural consistency and qualityEnsure all layers include robust verification mechanisms, including formal validation (where applicable), runtime checks, and auditabilityCollaborate with engineering, compliance, and client stakeholders to ensure the architecture meets regulatory, technical, and business requirementsMaintain a defensible architecture aligned with responsible AI principles, including transparency, traceability, and controlled capability exposureRequirements7+ years of experience in AI Engineering or a related roleStrong hands-on experience in Python development, including FastAPI, asyncio, pytest, type hints, and working within a monorepo structureExperience designing and implementing AI agent orchestration systems, including LLM gateways, tool gateways, streaming architectures (SSE and WebSocket), and agent-to-platform contractsUnderstanding of responsible AI governance, including capability boundaries, approval workflows, disclosure requirements, audit traceability, and regulatory control mappingExperience working with AWS cloud platforms in production environmentsExperience in rigorous testing practices, including unit, integration, and UAT/BDD testing, with enforced code coverage thresholds (≥90%)Proficiency with modern AI tooling, such as Cursor, MCP servers, prompt engineering techniques, and CLI automation tools (e.g., GitLab, Jira)Experience implementing observability using OpenTelemetry, Grafana or Phoenix, and structured logging practicesExcellent command of written and spoken English (B2+ level)Nice to haveUnderstanding of data architecture principles, including ontology design, taxonomy, conceptual/logical/physical modeling, data contracts, and gap analysisExperience designing and working with knowledge graphs in Neo4j, including dual-graph architectures, Cypher queries, graph sizing, versioning, and GraphRAG patternsAbility to design and implement domain-specific languages, including grammar definition, type systems, semantic validation, and execution pipelinesFamiliarity with security practices, such as PBAC/RBAC, JWT authentication, security context propagation, secret management, and secure code practices (SAST)Experience integrating cloud data platforms, including Snowflake and AWS services, such as EKS, IAM, and ALB, using Helm and GitOps tools like ArgoCDAbility to create and maintain architectural documentation using ADR/RFC processes and manage their lifecycle effectivelyKnowledge of formal methods, such as Alloy or TLA+ for high-assurance system designExperience with voice and multimodal AI systems, including STT/TTS pipelines using platforms like OpenAI or AWS BedrockUnderstanding of modern documentation practices (e.g., Diátaxis) and structured knowledge managementFamiliarity with compliance frameworks, regulatory traceability, and adversarial testing methodologiesExposure to additional ontology and modeling standards, such as RDF, SHACL, SKOS, or financial taxonomiesExperience with CI/CD quality gates, including tools like Bandit and Radon, and automated merge request review processesExperience building and productizing AI advisory solutions in regulated financial environmentsUnderstanding of end-to-end governance flows, including disclosure handling, capability restriction, human-in-the-loop approvals, and full audit chain implementation aligned with compliance requirementsWe offerWe gather like-minded people:Engineering community of industry professionalsFriendly team and enjoyable working environmentFlexible schedule and opportunity to work remotely within PolandChance to work abroad for up to 60 days annuallyBusiness-driven relocation opportunitiesWe provide growth opportunities:Outstanding career roadmapLeadership development, career advising, soft skills, and well-being programsCertification (GCP, Azure, AWS)Unlimited access to LinkedIn Learning, Get Abstract, Cloud GuruEnglish classesWe cover it all:Stable income (Employment Contract or B2B)Participation in the Employee Stock Purchase PlanBenefits package (health insurance, multisport, shopping vouchers)Strategically located offices featuring entertainment and relaxation zones, table tennis and football, free snacks, fantastic coffee, and moreReferral bonusesCorporate, social and well-being eventsPlease, note:The set of bonuses might vary based on the role you apply for – specifics will be discussed with our recruiter during the general interview.We will reach out to selected candidates exclusively.EPAM is a leading global provider of digital platform engineering and development services. We are committed to having a positive impact on our customers, our employees, and our communities. We embrace a dynamic and inclusive culture. Here you will collaborate with multi-national teams, contribute to a myriad of innovative projects that deliver the most creative and cutting-edge solutions, and have an opportunity to continuously learn and grow. No matter where you are located, you will join a dedicated, creative, and diverse community that will help you discover your fullest potential.