GenAI engineer in banking

The Senior GenAI Engineer is a core technical role within Ameriabank's formally established GenAI Enablement Service, operating under direct CIO sponsorship. This role focuses on designing, building, and deploying AI-powered capabilities that serve as shared, reusable infrastructure for the entire bank - AI agent workflows, retrieval-augmented pipelines, enterprise knowledge systems, and integrations with the bank's internal technology ecosystem. All development is conducted in a secure, on-premise environment where data sovereignty and regulatory compliance are non-negotiable.

• Design, develop, and deploy AI-powered automations, agent workflows, and RAG pipelines serving multiple business domains across the bank
• Design and evolve the Unified Enterprise Knowledge Base, including document ingestion, structured extraction, entity resolution, ontology/schema design, relationship extraction, evidence provenance, vector indexing, and hybrid GraphRAG retrieval.
• Develop and maintain reusable AI components, agent templates, and reference architectures that any tribe or squad can build on - embodying the 'build once, enable many' principle
• Design and implement Agentic AI frameworks - enabling autonomous multi-step workflows that coordinate across systems, tools, and processes
• Build and maintain integrations with enterprise systems - Jira, Confluence, TestRail, Azure DevOps - enabling AI-powered business and SDLC automation
• Develop internal APIs and microservices connecting AI components with bank infrastructure
• Integrate AI capabilities with the n8n (or similar) workflow automation platform - designing and maintaining production-grade AI-powered process automation pipelines
• Contribute to the SDLC AI Agent System - an end-to-end AI-assisted software development lifecycle covering requirements analysis, test generation, code review, and documentation
• Lead proof-of-concept development for new AI technologies - multimodal models, voice and speech processing, vision models, and Agentic AI frameworks
• Evaluate performance, accuracy, cost-efficiency, and security of models both in a controlled on-premise environment and in the cloud.
• Research and assess emerging open-source models and tooling - recommending adoption decisions based on evidence from structured evaluation
• Contribute to Armenian language model capability - evaluating multilingual models, fine-tuning approaches, and OCR solutions for Armenian document processing
• Ensure all AI solutions adhere to the bank's data classification framework - determining appropriate deployment tier (local on-premise, cloud SaaS, or cloud PaaS) based on data sensitivity
• Collaborate with Information Security, Cybersecurity, and Operational Risk to ensure secure data handling, audit trails, prompt management, and model transparency
• Document AI architectures, pipelines, experiments, and design decisions - contributing to the reusable pattern library and knowledge base of the GenAI Enablement Service
• Maintain the deployed AI capability registry - ensuring every production system has documented ownership, performance baselines, and governance records

• 5+ years in software engineering or applied AI/ML, with at least 2 years focused on GenAI systems in production environments
• Proven experience designing and maintaining RAG pipelines and vector databases.
• Experience with graph data modeling, entity/relationship extraction, entity resolution, graph traversal and query languages such as Cypher.
• Strong Python development skills, including testing, packaging, asynchronous programming, API development and production-quality software design
• Solid API development experience and microservice architecture; experience integrating AI systems with enterprise platforms (Jira, Confluence, or similar)
• Containerization (Docker), Linux systems, and GPU-based local deployments
• Experience with workflow automation platforms - n8n or similar - building production AI-powered process automation
• Understanding of data sensitivity classification, secure AI deployment practices, and compliance requirements in regulated environments
• Будет плюсом: Experience combining graph and vector retrieval for GraphRAG is highly desirable; Experience designing tool using LLM applications and multi-step AI workflows, including orchestration, state management, structured outputs, failure handling, human-in-the-loop controls and evaluation. Experience with LangGraph, LangChain, LlamaIndex or similar frameworks is desirable; on-premise AI environment experience preferred; Experience deploying and operating AI in secure, regulated industry environments - banking, financial services, or healthcare; Data engineering exposure - ETL pipelines, structured and unstructured data processing, SQL and Pandas proficiency; Experience collaborating within Agile or Agile@Scale delivery models; Contributions to open-source AI projects or internal innovation initiatives; Exposure to banking or financial services domain - understanding of core banking processes, regulatory environment, or financial data

• Условий в вакансии нет