Senior Software Engineer, AI/ML

 

Fishbowl is an industry leading, top supplier of manufacturing and warehouse management software for small, medium, and enterprise sized businesses across 40+ verticals. While our mission is to deliver amazing software, service, training, and support to our customers to help them grow and scale their business operations, our passion is helping people. Whether you are new to owning and operating a business, or you have been at it for 20+ years, Fishbowl provides simplicity and flow for business owners and makes it easier for them to focus on what they love most, running their business.

To support the mission of Fishbowl, we have recently partnered with Diversis Capital to invest in Fishbowl’s growth and market scale. We are well on our way to developing exciting new cloud-based products that will continue to surprise and delight our existing and future customers. We also have exciting plans to expand our efforts internationally and are focused on building a globally oriented team that will allow us to scale our operations and future market growth potential.

 

The Role

 

 

Fishbowl is seeking a Sr. Platform Software Engineer (AI/ML) to join our engineering team focused on building next-generation, AI-powered platform capabilities for our Inventory, Warehouse Management, and Manufacturing SaaS applications. This senior position offers a unique opportunity to work at the intersection of applied machine learning, intelligent planning systems, and scalable cloud architecture.

You will take the lead on technical initiatives involving LLM orchestration, machine learning, reinforcement learning, and decision systems that optimize resource allocation and inventory control across complex supply chain environments. Your work will be central to advancing our platform's intelligence, enabling customers to operate more efficiently, make predictive decisions, and adapt to rapidly changing conditions.

A core part of this role involves engineering context-efficient LLM systems to deliver AI augmented features to improve user efficiency and accuracy.  You'll also lead the implementation of Model Context Protocol (MCP) servers to enable AI agents to integrate and take actions on behalf of the user against our platform API’s.  Additionally, you'll design and optimize Retrieval-Augmented Generation (RAG) systems that allow agents to pull precise, relevant information into prompts without bloating token usage.

This is a high-impact role ideal for an AI engineer with SaaS platform experience who wants to own architecture, influence strategic direction, and drive innovation across a multi-product ecosystem. You will report directly to the Chief Architect and collaborate cross-functionally with product, cloud infrastructure, and development teams to build scalable, observable, and reusable AI systems in production.

Remote or hybrid work available (Orem, UT HQ). We emphasize outcomes over geography.

Responsibilities

 

 

•  LLM-Oriented System Design: Lead architecture of intelligent agent infrastructure that integrates LLMs into real-time SaaS workflows across manufacturing and inventory control.
•  Model Context Protocol (MCP): Design and implement MCP servers against our product API’s to enable user’s agents and application features to take action.
•  State-Driven Prompting (MDP-Style): Emulate Markov Decision Process patterns to reduce prompt bloat and improve agent determinism—updating agent state explicitly before generating each prompt.
•  Retrieval-Augmented Generation (RAG): Build and optimize RAG pipelines using vector search and semantic indexing to enrich prompts with highly relevant external knowledge while minimizing token overhead.
•  Orchestration and Prompt Control: Develop context managers, orchestrators, and prompt pipelines using tools like LangGraph, LangChain, or custom orchestration layers.
•  Agentic Workflow Architecture: Design and implement advanced agentic workflows using LLMs and orchestration platforms like N8N, enabling intelligent multi-step automation across system boundaries.
•  LLM + N8N Integration Patterns: Define and maintain standards for integrating LLM calls, MCP state management, and external system actions into reusable workflow nodes in N8N or equivalent orchestration tooling.
•  Workflow Extensibility and Governance: Establish secure, scalable patterns for authoring, deploying, and monitoring LLM-based agents within orchestration platforms—supporting reusability across product lines and customers.
•  Inventor