AI Engineer – Agentic Commerce, GenAI & Knowledge Graphs

Company: Fegmo

Experience: 3–6 years

Role: AI Engineer

Focus Areas: Agentic AI, LLMs, Agent Orchestration, Tool Calling, Knowledge Graphs, System Design

About Fegmo

Fegmo is building AI-powered solutions for the next generation of commerce. Our platform leverages AI agents to automate product data enrichment, catalog intelligence, content generation, product information management, and commerce workflows.

We are developing an agentic commerce platform where intelligent agents interact with enterprise systems, invoke tools, reason over product data, and use knowledge graphs to deliver accurate, contextual, and actionable outcomes.

We are looking for an AI Engineer with strong software engineering fundamentals, practical Generative AI experience, and a solid understanding of system design. You will work on building and orchestrating AI agents, developing reusable tools and services, and integrating knowledge graphs into intelligent commerce workflows.

Key Responsibilities

1. Agentic AI & Orchestration

• Design, develop, and orchestrate AI agents capable of executing multi-step commerce workflows.
• Build agent workflows involving planning, tool selection, execution, validation, and result synthesis.
• Develop reusable agent capabilities and tools that can be composed into larger workflows.
• Implement state management, context handling, task delegation, and coordination between agents where appropriate.
• Build reliable orchestration mechanisms with retries, timeouts, error recovery, and human-in-the-loop approvals.
• Evaluate when to use autonomous agents versus deterministic workflows to ensure reliability and control.

2. Generative AI & LLM Engineering

• Develop LLM-powered applications using modern language models, structured outputs, prompt engineering, and function/tool calling.
• Implement AI capabilities for product attribute enrichment, catalog classification, product understanding, content generation, and data normalization.
• Design context-building and retrieval mechanisms that provide agents with relevant product and business information.
• Evaluate model outputs for accuracy, consistency, groundedness, and business relevance.
• Optimize model selection, inference latency, token consumption, and operational costs.

3. Knowledge Graphs & Commerce Intelligence

• Design and implement knowledge graph-based representations of products, brands, categories, attributes, relationships, and commerce entities.
• Develop graph-based retrieval and reasoning capabilities to provide context to AI agents.
• Work with graph databases such as Neo4j and query languages such as Cypher.
• Integrate knowledge graphs with vector search, structured databases, and LLM-based applications.
• Build mechanisms to resolve entities, discover relationships, identify inconsistencies, and improve product intelligence.
• Contribute to the evolution of a reusable commerce knowledge layer that supports multiple agents and business workflows.

4. Tools, APIs & Enterprise Integrations

• Design reusable tools that agents can invoke through well-defined interfaces.
• Integrate internal services, external APIs, product information systems, search services, and data pipelines.
• Implement secure tool execution, input validation, permissions, structured responses, and failure handling.
• Explore and implement Model Context Protocol (MCP) where it improves interoperability between agents and tools.
• Build standardized interfaces that allow multiple agents to access shared capabilities without duplicating business logic.

5. System Design & Production Engineering

• Design modular, scalable, and maintainable AI services with clear service boundaries.
• Build asynchronous and event-driven workflows using queues, workers, and APIs where appropriate.
• Apply system design principles covering concurrency, idempotency, state management, caching, persistence, and fault tolerance.
• Develop automated tests, evaluation pipelines, observability, and performance monitoring for AI systems.
• Deploy and operate production services using cloud infrastructure and modern development practices.
• Collaborate with engineering and product teams to translate business requirements into robust technical solutions.

Required Qualifications

• 3-6 years of experience in software engineering, AI engineering, or machine learning engineering.
• Strong proficiency in Python, with experience writing clean, modular, testable production code.
• Hands-on experience building applications using LLMs, including tool/function calling and structured outputs.
• Good understanding of agentic AI concepts, multi-step workflows, and agent orchestration.
• Strong fundamentals in system design, APIs, databases, asynchronous processing, and distributed systems.
• Experience working with SQL databases and APIs; familiarity with search or vector retrieval systems.
• Understanding of data structures, algorithms, debugging, testing, and software design principles.
• Ability to reason about system reliability, AI quality, latency, cost, and architectural trade-offs.

Preferred Qualifications

• Experience with agent orchestration frameworks such as LangGraph, LangChain, or comparable technologies.
• Hands-on experience with Neo4j or another graph database, graph modeling, and graph queries.
• Familiarity with knowledge graphs, entity resolution, semantic search, embeddings, RAG, and hybrid retrieval.
• Experience building reusable tools, function-calling interfaces, or MCP servers and clients.
• Familiarity with message queues such as RabbitMQ, Kafka, Google Cloud Tasks, or AWS SQS.
• Experience with Docker, CI/CD, and cloud platforms such as GCP or AWS.
• Exposure to multimodal AI, computer vision, image generation, or product data intelligence.
• Experience at a strong product engineering organization or in a team with mature software engineering practices.

What Success Looks Like

• Build reliable, reusable agents that execute real commerce tasks rather than simply generate text.
• Develop an orchestration layer that coordinates tools, services, and agents with clear execution and failure-handling mechanisms.
• Enable agents to use the commerce knowledge graph for contextual retrieval and relationship-based reasoning.
• Improve the accuracy, consistency, and traceability of AI-generated product information.
• Deliver production-ready services with measurable quality, latency, reliability, and cost metrics.
• Contribute to a modular architecture that can support additional agents, tools, customers, and commerce use cases.

Ideal Candidate

We are looking for an engineer who combines strong coding ability with practical GenAI experience and a system design mindset. The candidate should be comfortable moving between LLM applications, agent orchestration, tool development, knowledge graphs, and backend architecture.

We value engineers who understand the fundamentals behind the frameworks they use, can evaluate architectural trade-offs, and can take ownership of a problem from design through deployment.

Experience at a large technology company is a plus, but demonstrated engineering quality, problem-solving ability, and production experience matter more than company names.

Interview Process

1 Coding and software engineering: Python, data structures, API design, and code quality.

2 Agentic AI: Design an agent workflow that invokes tools, manages state, handles failures, and validates results.

3 System design: Design a scalable orchestration service for commerce agents.

4 Knowledge graphs: Model products, attributes, brands, and relationships; explain how agents retrieve and use graph context.

5 Practical engineering: Implement or review a small agentic workflow with appropriate testing and error handling.