Backend Engineer for AI Agentic Systems
Project description
The paradigm of software is shifting from reactive to proactive. At Luxoft, we are at the forefront of this transformation, engineering the next generation of Agentic AI systems. We are seeking an exceptional Backend Engineer to architect the core infrastructure for software that doesn't just execute commands, but reasons, plans, and acts autonomously. As a key member of our elite AI Core team, you will move beyond traditional backend development to build the central nervous system for our AI agents. This role is for a true systems builder—someone passionate about creating robust, scalable, and intelligent platforms that empower autonomous AI to solve complex, real-world problems. You will be instrumental in crafting the bridge between raw data and digital cognition. What You'll Be Building: The Engine of Intelligence As our Backend Engineer for Agentic Systems, you will own the end-to-end lifecycle of the agent's core infrastructure—from its memory and logic to its deployment and operational resilience. - The Agent's Communication Backbone: You will wield FastAPI and Pydantic to design and build high-throughput, asynchronous, and self-documenting APIs. These are more than just endpoints; they are the high-bandwidth channels through which our agents perceive data, communicate their intent, and interact with external tools and systems. Your Pydantic models will define the rigorous data contracts that ensure absolute integrity across our distributed environment. - The Agent's Cognitive Framework: You will architect and implement the core systems that enable agentic behavior. This involves developing sophisticated state machines, implementing advanced reasoning loops (e.g., ReAct, Chain-of-Thought), and orchestrating the workflows that allow agents to plan, delegate tasks, and leverage a diverse suite of tools. - The Agent's Memory and Knowledge Base: An agent's power lies in its ability to learn and recall. You will be the steward of our agents' memory systems, leveraging the reliability of PostgreSQL. You will design and manage complex database schemas for storing conversation histories, learned knowledge graphs, and operational logs. Your expertise in database management and optimization will be critical to ensuring our agents can access information instantaneously as their knowledge scales. - The Agent's Operational Habitat: Our AI agents must be deployed with unwavering reliability and scale on demand. You will master Docker to containerize our entire agentic ecosystem, creating reproducible, isolated, and portable environments that guarantee consistency from development to production.
Responsibilities
- - Architect & Develop: Design, build, and maintain high-performance backend services in Python, using FastAPI for our core agentic platform. - Model Reality: Implement complex data validation and serialization schemas with Pydantic to ensure data integrity across our distributed systems. - Empower Agents: Develop the core logic for AI agents, including tool integration, state management, and multi-step reasoning frameworks. - Manage Data Persistence: Design, optimize, and manage PostgreSQL databases. Write efficient SQL queries, manage migrations, and ensure the long-term health and scalability of our data stores. - Containerize & Deploy: Create and maintain clean, efficient Dockerfiles and Docker Compose configurations for all backend services, streamlining our deployment pipelines. - Collaborate & Innovate: Work in close partnership with AI Researchers, Prompt Engineers, and Frontend Developers to deliver cohesive, end-to-end intelligent solutions. - Test & Harden: Author comprehensive unit and integration tests to guarantee the reliability and robustness of the foundational systems powering our agents.
SKILLS
Must have
- Technical Excellence: o Proven Professional Experience: A minimum of 3 years in backend engineering, with a strong portfolio of building scalable, production-grade systems using robust Python skills. This includes deep, hands-on experience with database management. o Deep expertise in FastAPI and a strong appreciation for its asynchronous capabilities. o Mastery of Pydantic for creating strict, type-safe data models that enforce architectural clarity. o Experince in any of common agentic frameworks such as langGraph, Pydantic-AI, Google ADK, Agno or any similar framework o Strong command of PostgreSQL, including advanced schema design, query optimization, and indexing strategies. o Production-level experience with Docker, with a proven ability to write clean, multi-stage Dockerfiles and manage containerized applications. - The Agentic Mindset: o A profound passion for AI and a demonstrable interest in agentic systems. You have likely experimented with frameworks like LangChain, LlamaIndex, or AutoGen, or you possess a strong desire to build such systems from fundamental principles. o A systems-thinking approach. You can visualize how disparate components—LLMs, databases, APIs, and tools—integrate to form a single, intelligent entity. - Personal Attributes: o Insatiable Curiosity: You are constantly learning and exploring the bleeding edge of AI and software engineering. o Ownership Mentality: You don't just write code; you take ownership of the systems you build and are accountable for their performance and reliability. o Collaborative Spirit: You thrive in a team environment, sharing knowledge and elevating those around you.
Nice to have
- Experience with vector databases (e.g., Pinecone, Chroma, Weaviate). - Knowledge of cloud platforms (AWS, GCP, Azure) and infrastructure-as-code (Terraform) - Familiarity with CI/CD pipelines (e.g., GitHub Actions, GitLab CI) - Experience with message queues (e.g., RabbitMQ, Kafka) for asynchronous task processing.
