AI Product Engineer | REMOTE

Job description

**Powering Trade For The Emerging Cannabis Industry** Distru is the [#1 ERP in the Cannabis Industry](https://www.g2.com/products/distru/reviews) with $3B in annual sales processed through our platform! We are a fully remote and global team building a software platform to streamline the supply chain for the cannabis industry. Most of the top cannabis brands and distributors across the United States use our software to manage their entire operation and stay compliant. As part of our next phase of growth, last year we launched Distru Commerce; a wholesale platform that extends our ERP into the B2B marketplace layer, giving operators a single system to manage their internal operations and their external wholesale sales channels in one fully integrated system. But we're not stopping here. AI is now a core part of how we build and where we're placing our bets. The AI Order Agent is live and already changing how our customers operate: buyers can submit orders from any source and any format saving time spent on manual data entry. That's one piece of a bigger picture: using AI to take the repetitive, heavy-lifting work out of running an ERP so operators can focus on running their business. We take a tremendous amount of pride in our culture and have [5/5 stars on Glassdoor](https://www.glassdoor.com/Reviews/Distru-Reviews-E2957092.htm) because we truly care about our people. We are stable, profitable, and heavily investing in growing new product lines to expand our reach. Our investors include [Felicis Ventures](https://www.felicis.com/), [Global Founders Capital](https://www.globalfounders.vc/), [Poseidon Asset Management](https://www.poseidonassetmanagement.com/), and [Village Global](https://www.villageglobal.vc/) who have funded companies such as Shopify, Cruise, HelloFresh, Facebook, Slack, and Opendoor. **The Opportunity** We're looking for an AI Product Engineer, working directly with our CTO, to architect and build an agentic harness — think of it as the Claude Cowork for vertical-specific SaaS. This product will bridge the tools businesses already use to run their operations (ERP, CRM, spreadsheets, calendars, email, analytics, compliance systems) and fundamentally change how they interface with their own data and systems. This is not a "bolt AI onto existing features" role. You'll be building a new, standalone agentic product from the ground up, which means you'll operate as both the engineer and the product owner: scoping the problem, making architecture calls, and shipping without a lot of structure around you. You'll report directly to the CTO and work closely with product, engineering, and design as this fits into Distru's broader AI-native roadmap. Distru is at an inflection point. We have deep market penetration in cannabis ERP, a commerce layer live and growing, and a roadmap that's increasingly AI-native. This role owns one of the most ambitious pieces of that roadmap — building the connective tissue that lets an entire industry run its business through an agentic interface. **Our Team Culture** Rather than us defining our culture, [check out what the team has to say](https://www.glassdoor.com/Reviews/Distru-Reviews-E2957092.htm)! ### What You'll Own **Build the agentic harness** - Architect and build a multi-agent system in production — one capable of safely operating across a business's core tools (ERP, CRM, spreadsheets, calendars, email, analytics, compliance systems) - Build and maintain MCPs (Model Context Protocol integrations) that connect the harness to real business systems - Own production system concerns end-to-end: reliability, observability, permissions, and safe workflow execution — this is a system that will take real actions on customers' data, and the bar for getting it right is high - Tech Stack:LangGraph, Next.js, React, Prisma, Claude Code **Own the product, not just the code** - Since there's no existing playbook here, you'll define scope, make the call on tradeoffs, and drive execution with minimal oversight - Design for the user, not just the architecture — a product-first mindset that prioritizes how this actually gets used over what's technically interesting to build - Proactively identify and raise risks or open questions before they become blockers, and bring structured solutions, not just problems **Stay at the edge of the AI ecosystem** - Keep pace with a fast-moving space — agent frameworks, harness toolkits, MCP standards — and bring in better approaches as they emerge rather than waiting to be told - Work hands-on with tools like Claude Code as part of your own build process ### Who You Are **Must-Haves** - 5+ years of full-stack development experience, ideally in B2B SaaS applications - Hands-on experience building multi-agent harnesses in production — not just prototypes or personal projects - Experience building MCPs (Model Context Protocol) that are actually used in production - Strong grasp of production system concerns: reliability, observability, permissions, and safe workflow execution - Experience across product and engineering — you've operated as both. You should be comfortable owning scoping and decision-making, not just implementation - Proficiency in modern technologies such as React, TypeScript, Node, PostgreSQL, GraphQL, and LangGraph, with the ability to quickly pick up new tools and frameworks - Startup agility: experience at an early-stage startup (under 20 engineers) where you've operated effectively in ambiguous, unstructured environments - Low ego, team-first mentality — you measure success by what shipped and whether it worked, not by internal politics - Clear, proactive communicator — you raise concerns and propose solutions before being asked **Nice-to-Haves** - Background building in supply chain, ERP, CRM, B2B marketplaces, or accounting software - Experience with UX design - Familiarity with agentic systems and multi-agent orchestration beyond what's listed above (e.g., evaluation frameworks, agent-to-agent protocols) - Experience using Claude Code or similar tools to build products independently - Cannabis industry experience