AI & Agentic Systems Architect İş İlanı

Rakita is an AI-first technology company built on over two decades of enterprise software expertise and the experience gained through 250+ successful projects across 10 industries. Born from PATH's extensive experience in software development, Rakita focuses on transforming how businesses operate through intelligent, scalable, and purpose-built AI solutions. Our growing ecosystem of AI-powered products spans software development, recruitment, human resources, customer support, marketing, and business operations. By combining deep technical expertise with a product-driven mindset, we build solutions that go beyond traditional automation, enabling businesses to work smarter, faster, and more efficiently. Driven by our vision of making AI an integral part of everyday business operations, we are building the next generation of intelligent enterprise solutions. We are looking for an AI & Agentic Systems Architect to join the Rakita team. Requirements Strong experience in software architecture and hands-on backend or platform engineering. Practical experience designing, implementing, and operating LLM-based applications and tool-using agents. Understanding of agent planning, orchestration, context management, structured outputs, retrieval, and memory. Experience with distributed systems, APIs, event-driven architecture, asynchronous processing, and durable workflows. Ability to design safe execution of agent-generated code and commands using sandboxing and controlled environments. Experience with containers, Kubernetes, cloud infrastructure, CI/CD, and infrastructure as code. Understanding of AI evaluation, failure analysis, observability, and cost optimization. Strong knowledge of identity, authorization, secrets management, and application security. Ability to write production-quality code, review implementations, and communicate architectural decisions clearly. Evidence of delivering working systems, with clear ownership of architecture and implementation. Preferred qualifications Strong C#/.NET experience, alongside Python for AI development. Experience with Azure, Azure OpenAI, .NET Aspire, and enterprise AI integrations. Experience with agent frameworks and workflow engines, with the ability to assess their limitations and build custom components where needed. Familiarity with Model Context Protocol (MCP), agent interoperability, and enterprise tool integration. Experience with GitLab or GitHub, internal developer platforms such as Backstage, and automated software delivery. Experience with PostgreSQL/pgvector, vector search, hybrid retrieval, and enterprise RAG. Experience designing multi-tenant SaaS platforms and dynamically configurable systems. Familiarity with DDD, platform engineering, developer experience, and sociotechnical architecture. Experience in retail, e-commerce, or enterprise environments with complex integrations. Key responsibilities Design agentic architectures: Define agent responsibilities, interaction patterns, tool interfaces, memory, state management, and appropriate boundaries of autonomy. Build autonomous agents: Implement agents that plan, execute actions, use tools, validate results, and recover from failures within defined permissions. Develop orchestration: Build workflows supporting task delegation, parallel execution, durable state, retries, timeouts, cancellation, checkpoints, and human approvals. Engineer agent infrastructure: Design isolated execution environments, containerized workers, queues, scheduling, resource limits, secrets management, and controlled network access. Build agent harnesses: Establish runtime controls for context management, tool execution, output validation, execution budgets, tracing, and reproducibility. Enable AI-native SDLC: Integrate agents with issue tracking, source control, CI/CD, testing, security scanning, documentation, and deployment workflows. Integrate enterprise knowledge: Design retrieval and memory systems using code repositories, architecture standards, service catalogs, and operational documentation. Embed security: Implement least-privilege access, tenant isolation, auditability, prompt-injection defenses, and safeguards against unauthorized actions or data exposure. Establish evaluation and observability: Measure task success, output quality, tool-use accuracy, reliability, latency, and cost through repeatable evaluations and production monitoring. Lead architectural decisions: Evaluate models and frameworks, document trade-offs, create reference implementations, and guide teams toward reusable patterns. Drive adoption: Collaborate with developers, platform engineers, security specialists, and business stakeholders to turn suitable use cases into measurable outcomes.