Maria at ReturnStack requested a call to discuss an area where StoryHouse can provide support, centered on late-stage Amazon deal due diligence. Amazon requires three years of audited financial statements, which ReturnStack does not have at under one year old; the alternative documentation package includes a cap table copy, bank statements, and an investor letter of support. Amazon pays in 60 days, so working-capital requirements run close to the current bank balance. The deal meeting was set for Friday 3pm.

Investor letter of support: Miles to draft with Alexei, whose retail background adds credibility. Purpose is to demonstrate financial stability for a multi-year contract, messaging that StoryHouse is excited about the Amazon deal and can help ReturnStack access additional capital for rapid expansion. Maria to send specific Amazon language and requirements by text.

Industrial real estate need: expansion required to one of four specific US locations, warehouse minimum 30,000 sq ft, operational by July on a tight timeline. Miles has a portfolio company with an industrial real-estate customer network and will also check LP connections; Maria to send a forwardable blurb for outreach.

Next call with ReturnStack to further diligence, ask questions, and request introductions to wrap up DD. Internal sense: offer $800K at a $16M post to get 5%, or go the time-bound discount route. Very positive call, more clarity on Amazon and current pilot partners.

Amazon partnership: Amazon needs ReturnStack short-term to automate returns processing and is saving ~$1B on returns this year through structural changes. Focus is high-value items ($100+) with fraud risk; Amazon wants to bypass its warehouse for predicted recycling/liquidation items and needs fraud detection outside its warehouses. ReturnStack provides fraud evidence for retro-charging customers. The current program bypasses 14-15% of low-ASP items directly to liquidator/recycler. RFP starts January, decision by February, annual contract, competing against manual sortation providers, with ReturnStack the only provider offering fraud detection. Pricing $1.50 versus Amazon’s internal $3-3.50 cost.

Current pilot, Charming Charlie and Redo: Redo partnership launched last week, processing returns for Charming Charlie ($60-70 ASP, post-bankruptcy and now online-focused) with a focus on item matching and data creation. Redo acts as the returns portal and ships units to ReturnStack; an NDA was signed for additional customers via Redo, with volume commitments expected within 1-2 weeks.

Financial model, three revenue sources: direct partnerships (Phoenix and others) at $3/unit, split ~70% basic fraud detection and 30% full service including resale; the Amazon partnership at $1.50/unit starting at 0.1% of ~1B annual returns; and direct retailer partnerships targeting $4/unit by 2027, aiming for 3-5 direct partnerships by end of 2026. 2025 target of 1M units processed, automating 60-70% of assessment questions at 95%+ accuracy.

Go-to-market: ~100-customer pipeline from Amazon connections, 10 demos completed including REI, Lululemon, and early conversations with others; conference strategy around Deliver and NRF. Competition: a ReturnPro CEO conversation the prior day surfaced a 150-person product team with bureaucratic and legacy-integration challenges at $6 pricing versus ReturnStack’s $4 target; Optoro faces similar manual-process constraints. Team: 50/50 equity split between Mayank and Maria, both geographically flexible; needs two additional engineers (~$200-250K each). Funding: raising a $350-700K bridge ($150-200K for warehouse expansion, remainder for AI talent and API integrations), 3-4 months runway, with a larger institutional round planned for Q1 and one VC in DD (final meeting pushed to January). Next steps: reference calls with Redo’s chief of staff and professional references, and continued pricing discussion for the bridge.

Team background: Maria, Harvey Mudd, PhD computer vision (UCLA), Snapchat engineer #30 who started the AR-filters research team, tech M&A due diligence and product launches for $100M+ acquisitions, five years at Amazon leading 50-60 person teams, recently quit Uber for ReturnStack full-time. Mayank, 12 years at Amazon: grocery from four walls to $5B across 60 cities, Amazon Flex last-mile from 300M to 5B packages/year leading an 800-person team, and returns processing handling ~1B units/year.

Market problem: return rates doubled in 4-5 years (8-9% to 16% overall), apparel returns 20-25% (luxury up to 25%), ~$900B of inventory returned in the US last year with only $300-400B recovered, so two-thirds of return value is lost to inefficiency. About 80% of customers consider free returns a “birthright,” and roughly 50% of returns currently go straight to landfill because manual processing makes low-value items uneconomical.

Solution: AI-powered returns processing using computer vision across a three-stage flow (customer return portal integration, warehouse processing with fraud detection, automated resale listing). Core tech includes image analysis comparing returned items to the product catalog, fraud detection cited as 3x better than competitors, automated condition assessment and resale pricing, and a claimed 40% labor cost reduction (~$4/unit savings). Operations started in Mayank’s garage buying liquidation inventory and now run from an Indianapolis warehouse, with an industry partner pilot processing real returns, an Amazon pilot starting in December focused on fraud detection, and a third partner interested for January.

Fraud: 10-15% of returns are fraudulent (3-5% wrong-item swaps, the rest abuse/overuse); the system flags suspicious returns with photographic evidence and lets retailers configure automatic denial versus manual review, with customer risk tracked across retailers. Business model: two streams, a $4/unit processing fee (versus ReturnPro’s $6) and 10% resale revenue share (versus 15-20% for competitors); target customers are $100-500M revenue retailers and D2C brands with ASP above $50; ~70% of returns processing is currently outsourced, and ReturnStack partners with returns-portal providers such as Get Redo for volume.

Competitive landscape: first-stage portals include Narvar and Get Redo; second-stage processing has limited players in Amazon, ReturnPro, and Optoro. ReturnPro is a 20-year-old legacy company at ~$500M revenue with Walmart as a customer, running pen-and-paper processes at higher pricing; Amazon processes efficiently but does not offer external services. Tech advantages: fine-tuning multimodal LLMs for defect detection, a proprietary product database, Shopify API integration for real-time catalog access, and fraud detection Amazon currently lacks. Scaling: company-owned initial warehouses for control, later a “business in a box” model for warehouse partners, with category expansion from apparel/footwear (35% of returns) into electronics. Funding: immediate need of $350-700K for warehouse expansion on a two-week timeline, uncapped SAFE with 20% discount preferred, Q1 institutional round targeting $10M with pre-term-sheet diligence already underway; StoryHouse interest in leading a smaller bridge round with a cap, adding value through retailer-network introductions (Alexei). Next steps: send the DD questionnaire, review the financial model, explore an advisory/operating arrangement with Alexei, and a two-week timeline for the bridge decision.