# Moxie Retail, a Genzeon company — Retail AI Solution Architect

- Generated: 2026-08-31 11:12:02 AM EDT
- Original/canonical posting: https://www.linkedin.com/jobs/view/4459216513/
- User-provided LinkedIn posting: https://www.linkedin.com/jobs/view/4459216513/
- Posting time: approximately 2026-08-31 10:12 AM EDT (estimated from LinkedIn's '1 hour ago' at capture)
- Applicants: Be among the first 25 applicants
- Work model/location: Remote within the United States with roughly 50% travel to client sites; LinkedIn location: United States
- Compensation: Not disclosed
- Travel: Roughly 50%
- Positioning track: Technical IC
- Fit outcome: FAIL — 47%

## Direct-match strengths

Agentic AI, Python/FastAPI, RAG, MCP, vector search, APIs, graphs/ontologies, executive presentations, and hands-on AWS delivery.

## Hard or material gaps

Hard failures: deep retail-operations experience is explicitly required and unsupported; roughly 50% travel exceeds the 5% maximum; pursuit, proposals, SOWs, pricing, margin, and partner co-selling are central.

## Evidence map

1. Agentic AI engineering — direct production LLM, RAG, agents, orchestration, MCP, evaluation, and workflow evidence.
2. Client-facing leadership — direct executive advisory and technical presentations to 200+ AWS customers.
3. Data and integration — direct SQL, APIs, event-driven systems, ERP-data integration, Docker, and cloud evidence.
4. Graph and ontology work — direct Neptune, SPARQL, UMLS, and knowledge-system evidence.
5. Retail operations — unsupported mandatory expertise in order lifecycle, allocation, ATP, BOPIS, replenishment, and retail platforms.
6. Travel and commercial pursuit — 50% travel conflicts with Keith's maximum; pricing/margin/SOW/co-sell ownership is outside target scope.

## Full normalized job description

Retail AI Solution Architect Forward Deployed Pre-sales discovery, solution design, and hands-on delivery for agentic AI in retail operations. Location: United States, remote — with roughly 50% travel to client sites Level: Director / Senior Director · typically 8+ years, at least 4 of them in retail Reports to: General Manager, Moxie Systems · Retail Business Solutions practice Type: Full-time Why this role exists Moxie Systems works only in retail. We serve four verticals — Fashion & Apparel, Lifestyle & Beauty, Convenience & Fuel, and Luxury — and our clients are largely past the pilot stage. They have run the demos. What they need now is someone who can look at their actual order, inventory, and product data and say honestly what agentic automation will and will not do for them. That conversation is the front end of this job. You lead discovery, turn what you hear into a solution design and a phased plan, then stay attached as the technical anchor while the first phase gets built. You are in the room before the statement of work is signed, and still there when the first agent runs against production data. We call it Forward Deployed because you deploy with the client, not behind them. What this role is not: A sales engineer demoing a fixed product, an architect who draws a diagram and hands it off, or a builder waiting on a requirements document. All three exist elsewhere at Moxie. This role is the seam between them. Where the bar actually sits The combination of skills below is uncommon, so we would rather be explicit than have you self-select out of a role you would be strong in. Two dimensions are firm. The rest are a menu, not a checklist. Retail domain depth: Required You have lived inside retail operations. You know what breaks, and why the last project failed. Client-facing presence: Required You can run a room of skeptical architects and executives, and you build your own materials. Hands-on AI engineering: Calibrated Prototype and critique real code. You need not have been a full-time production ML engineer. Data and integration: Calibrated Real depth in one area, working literacy across the rest. See the menu below. Frontend craft: Not required A demo that holds up on a screen share is enough. Production UI sits with our engineers. Consulting pedigree: Not required We hire for judgment, not for a logo or a certification. How the role splits: 35%- Pursuit — discovery, solutioning, proposals, executive presentation 25%- Solution architecture and technical design 25%- Hands-on build — proofs of concept and first-phase delivery 15%- Practice building — reusable assets, partner enablement, mentoring Environment AI and agents: Anthropic SDK, OpenAI SDK, MCP, agentic orchestration, evaluation tooling. Backend: Python, FastAPI, async patterns. Data: PostgreSQL, Neo4j and Cypher, vector stores, client OMS and ERP schemas. Platforms: UiPath, Microsoft Azure, client commerce and OMS platforms. Delivery: Docker, Git, cloud and on-premise. Working tools: Slack, Confluence, PowerPoint and Word. What you need Two things are required. Everything below them is a menu, not a checklist. Strong candidates are deep in two or three of these areas and conversant in the rest — tell us which are yours. Retail domain depth Required You cannot lead a discovery session on agentic readiness while learning what available-to-promise means. Fluency across the order lifecycle and distributed order management, sourcing and allocation, ATP, returns and reverse logistics, store fulfillment and BOPIS, and replenishment — with real depth in at least one of our four verticals. Client-facing leadership Required You have run structured discovery with mixed IT and business audiences and surfaced the constraint the client did not plan to mention. You present to VP and C-level stakeholders, whiteboard under challenge, and build your own slides. You have contributed materially to SOWs and proposals, and can build a business case a finance leader will not dismantle. Retail platforms Hands-on with one or more of: Manhattan Active, Blue Yonder or legacy JDA, Oracle Retail or Oracle Cloud ERP, SAP, Salesforce Commerce Cloud, Shopify Plus, NetSuite, or a comparable enterprise OMS, WMS, or commerce platform. Retail data Any of: PIM and MDM work (Stibo STEP, Informatica, or similar); PCI and PII constraints; EDI; POS or loyalty data; the integration debt inside a twenty-year-old estate. Applied AI Credible hands-on work in one or more of: agentic tool-calling loops and multi-step chains (Anthropic SDK, LangChain, or equivalent); RAG and vector search; evaluation and hallucination diagnosis; cost and latency tuning. Data and integration Depth in at least one of: SQL and data modeling against unfamiliar OMS or ERP schemas; REST and API design, auth, rate limiting, event-driven and batch patterns; deployment with Docker across cloud and on-premise. Differentiators None required, and any one stands out: Neo4j, Cypher, or graph and ontology modeling; UiPath agentic automation; Azure AI or Fabric and partner co-selling; IBM i and AS/400 modernization; prior forward-deployed or resident architect work. What you will do Pursuit and discovery. Lead discovery with client architects, operators, and executives. Qualify honestly, including telling us when an opportunity is not real yet. Author solution narratives, proposals, and POC scopes with explicit success criteria, and defend the design under pressure. Partner with UiPath and Microsoft field teams on joint accounts. Solution architecture. Translate findings into reference architectures, integration designs, and phased plans. Document the real trade-offs — graph versus relational versus vector, agentic versus deterministic, buy versus build. Size and price the work with delivery leadership so both the estimate and the margin survive the project. Name data readiness constraints early, in writing. Hands-on delivery. Build proofs of concept yourself, against client data, in client environments. Anchor the first delivery phase, then hand off cleanly. Hold the architectural line as the solution scales, and stay close enough to production to know whether what you sold is actually working.

## Artifact metadata

- Resume and cover letter: Not generated under the fit gate.
- Archived JD capture: https://files.keithsteward.com/Moxie_Retail/Keith_Steward_Retail_AI_Solution_Architect_JD_capture.md
- Google Drive used: No