The Shop With Me Year 2 Roadmap — What We're Building, What We're Not

Most company roadmaps are marketing artifacts. This one isn't trying to be. It's the actual list, with the actual priority, and with the things we're explicitly not building included so the priorities make sense. If you're a current or prospective customer, this is what you can plan around. If you're an investor or advisor, this is what we're spending year-two effort on.

The tldr: deepen rather than broaden. Five focused areas. We are not adding categories.

What's shipping in year two

1. Bilingual support beyond English and Spanish

Today Remi ships in English and Spanish, and that covers the majority of our pilot footprint in Southern California. Year two we add three more language modes, prioritized by where our existing customers and inbound interest are concentrated:

• Vietnamese — significant indie-retail presence in California and Texas; underserved by existing AI products.
• Korean — same story, with concentration in LA and Atlanta metros.
• Tagalog — strong inbound interest from convenience operators in Hawaii and parts of California.

These are not just translation work. Each language requires separate voice model selection, separate tone calibration with native-speaker advisors, and separate testing in real store environments. We're committing to ship one of these by end of Q2 and a second by end of Q4. The third moves into year three if the data says it should.

We're not adding French, Mandarin, or Arabic in year two. They're on the longer list and they'll come. The principle is one language per quarter at most, done well, instead of ten translated badly.

2. Deeper POS integrations

We integrate with two POS systems today, with a third in late-stage work. Year-two target is four to six mainstream POS integrations, with bidirectional inventory sync where the POS supports it.

The priority order is set by what our existing pipeline asks for:

1. Clover — broad install base in our target verticals.
2. Square for Retail — POS AI features don't replace floor concierge, see our comparison post.
3. Lightspeed Retail — strong fit for liquor and beverage stores.
4. Cumulus / RetailEdge — common in the older indie-retail tier our pilot stores came from.

Bidirectional means: when a sale happens at the POS, inventory in Remi's catalog reflects it within minutes; when a store owner adjusts a product description in our dashboard, it can optionally write back to the POS catalog. Today most of our integration is one-way (POS → Remi). Two-way is harder, requires more vendor cooperation, and is what the customer actually wants.

We are not building a POS. We will not build a POS. The integrations are how we play with the systems stores already pay for.

3. Mobile companion app

The mobile app is built. It's been built for months. It's sitting in App Store and Google Play submission queues with the kind of small process problems that always slow consumer-app launches — privacy review, screenshot revisions, store listing copy. By end of Q1 it's live on both stores.

What the app does for the shopper:

• Browse the kiosk-store inventory of any participating Shop With Me store.
• Save favorites and get notified when something restocks.
• Continue a conversation with Remi from the kiosk on their phone after they leave the store.
• Wallet-style identification at the kiosk, so a returning shopper gets context-aware recommendations.

What the app does for the store owner:

• Extends Remi's reach beyond the four walls.
• Builds an opt-in customer relationship the store actually owns (not platform-mediated).
• Closes the loop on attribution — a recommendation made at the kiosk that converts on the app is now traceable.

What the app deliberately does not do: dark-pattern engagement loops, location tracking when the app is closed, social-graph extraction. We held the design tight to consumer trust because indie retail customers will walk away from anything that feels invasive.

4. Multi-store admin maturity

Year one's multi-store support was MVP. The schema supports multiple stores per customer account; the dashboard supports switching between them; and a few of our customers have two locations and use it.

Year two we make this real:

• Cross-store reporting. A multi-store owner sees aggregate analytics across all their locations, with drill-down to any one store.
• Catalog templates. A two-store operator can define a base catalog and have it applied to both stores with per-location pricing and stock overrides, instead of editing each store separately.
• Role-based access. A regional manager gets read-only across all stores; a store-level clerk gets edit access on their location only. Today everything is owner-level.
• Bulk operations. Mass price changes, mass product deactivation, scheduled rollouts — the table-stakes admin operations a 5+ location operator needs.

This is a year-long project, not a one-quarter feature. The underlying data model already supports it; the work is UI, permissions, and the testing matrix that comes with multi-tenant access control.

We are not opening a multi-store tier with full enterprise procurement (SOC 2, MSA templates, dedicated CSM) in year two. That's year three. The multi-store tier we ship is for owner-operators with 2-20 locations, not for chains with corporate IT.

5. Opt-in surveillance integration

This is the most carefully phrased item on the roadmap. We have a surveillance service stub in the codebase. It is not running. We will not turn it on without three preconditions met:

1. Opt-in by store owner, defaulted off. Never silently enabled.
2. Opt-in by shopper where local law requires it. California and Illinois set the bar. We meet it.
3. No facial recognition without separate, explicit consent. AWS Rekognition is wired but disabled. Even when enabled, it operates only on shoppers who've affirmatively enrolled.

What surveillance integration is for: shrinkage signal for the store owner, queue-length awareness so Remi can adjust her response time, and aggregate foot-traffic analytics that help with merchandising. What it isn't for: covert biometric capture, identity tracking, or any feature that fails our great-grandfather test.

If we cannot ship this with the privacy model intact, we will not ship it. The roadmap commitment is to do the work, not to ship at any cost.

What we're explicitly not building

A roadmap without anti-goals is just a wishlist. The things we have decided not to do in year two:

• Enterprise sales motion. We are not hiring an enterprise AE in year two. The product isn't ready for it and the deal cycle would distort the roadmap.
• White-label or reseller program. Several inbound asks, none of them good enough yet to displace direct customers.
• Hardware vending or proprietary silicon. We will ship a designed enclosure (year-one retrospective covers why), but the kiosk itself stays on commodity tablets.
• Restaurant ordering kiosks. Different vertical, different UX, different competitive set. Not our fight.
• Generative product imagery / video. Cool demo, not a customer ask. Skipping.
• A storefront marketplace. We are not Shopify. We integrate with the platforms our customers already use. We don't run a marketplace.
• Crypto payments. Asked occasionally. Real customer demand: zero. Skipping.

The list of things we're not doing is longer than the list of things we are. That's intentional. Discipline at this stage is the difference between a focused product and a rollercoaster.

How we sequence within the year

Roughly:

• Q1: Mobile app launches on both stores. First non-EN/ES language ships. Clover integration in beta.
• Q2: Square for Retail integration ships. Multi-store cross-store reporting v1. Hardware enclosure pilot in three stores.
• Q3: Second non-EN/ES language. Lightspeed integration. Multi-store catalog templates and role-based access.
• Q4: Surveillance opt-in goes live in pilot stores with full consent flow. Hardware enclosure GA. Fourth POS integration.

These dates will move. Roadmaps always slip in the direction of more focus and fewer features. We will publish revised dates if the slip is more than a quarter.

What this means for current customers

If you're already a Shop With Me customer, all of this lands in your dashboard automatically. No version upgrades, no migration headaches, no new contract. The pricing tier you're on today carries forward. New tiers (multi-store specifically) will be additive — existing single-store customers stay where they are unless they choose to upgrade.

If you're considering becoming a customer, the question to ask is whether the year-two roadmap matches what you'd want from this product over the next twelve months. If it does, book a demo. If there's something missing from this list that's a deal-breaker for you, tell us. The roadmap is set, but it's set based on customer evidence, and the right kind of feedback can move things up the priority order.

Frequently asked

Will any of this be available in beta before GA? Yes. Existing customers get early access to integrations and the multi-store admin work. We pick beta participants based on fit, not first-come.

Is the roadmap available to investors in more detail? Yes — there's a more detailed internal version with engineering estimates and dependencies. We share it under NDA with investors and advisors who are actively in conversation with us.

What about facial auth? The CLAUDE.md mentions it. It's wired but disabled. It will remain disabled until we have the consent and privacy model right. If that takes longer than year two, so be it.

Why isn't enterprise on the roadmap? Because we'd do it badly today and erode trust with both the enterprise customer and the indie customer at the same time. We'll get there. Not in year two.

What if I want a feature that isn't on this list? Tell us. The roadmap is set but features get bumped up when the customer evidence is strong. Email or request a demo and walk us through the use case. We listen, even when we say no.