Prepared by UNO | April 25, 2026 For: 9 / Owner review Status: Frank capability-vs-gap analysis for Q3 2026 launch decision
9enterprises has a real AI communications backbone — voice, text, email, Telegram, Twilio, ElevenLabs, Supabase, Stripe credentials in the Locker — but none of it is wired to Agent9 the product. What exists is internal infrastructure for Jasson's personal AI operations, not a multi-tenant real estate platform. The gap between what we have and what Agent9 needs to launch in Q3 2026 is not a polish gap. It is a full product-build gap: we need seller onboarding, KYC/ID verification, MLS submission via an Ohio broker partner, lockbox hardware integration, offer and document generation, e-signature, escrow coordination, a buyer-facing search layer, Stripe Connect payouts, and a mobile-friendly front end — none of which exists today. A 2-person team cannot ship all of that to production quality in 22 weeks. The honest Q3 2026 answer is: a demo MVP with 2-3 real transactions is achievable by Q3; a production platform capable of volume Ohio transactions is a Q4 2026 or Q1 2027 story. Q3 can and should be the demo launch date — but the Owner needs to know that Q3 "launch" means a live pilot with wizard-of-oz back-office, not a scaled platform.
| Component | Status | What it actually does |
|---|---|---|
| comms-hub.mjs (port 3457) | LIVE | Internal 9 messaging router. Telegram, iMessage, email (Gmail SMTP), voice inbound. Relay + autonomous mode. Not multi-tenant. Not customer-facing. |
| voice-server.mjs (port 3456) | LIVE | Twilio webhook + ElevenLabs TTS. Single-persona voice AI for Jasson's calls. Ephemeral cloudflared tunnel — Twilio webhook URL changes on restart. Not production-stable for customer calls. |
| ElevenLabs | LIVE | eleven_turbo_v2_5, Laura voice. ~$0.30/min at turbo pricing. Sub-second latency in practice. Fully usable for Agent9 buyer call handling with a new voice persona. |
| Twilio | LIVE | SIP-trunked numbers, SMS bridge, voice. A2P 10DLC registration status unknown — if not registered, marketing SMS is blocked at carrier level. This is a gap to verify before launch. |
| cloudflared tunnel | LIVE | Ephemeral URL. Must be replaced with a named Cloudflare Tunnel for production voice handling — ephemeral URLs are not acceptable for Twilio webhook config in a product. |
| Component | Status | Tables / Notes |
|---|---|---|
| SQLite (data/9-memory.db) | LIVE | 8 tables: messages, actions, authority, memory, tasks, decisions, health_events, audit_log. Single-Mac, no automated backup. Not suitable as a product database without migration. |
| Supabase | LIVE | Mirrors messages and actions from comms-hub. Anon key in Locker. Has real-time capability. This IS the right cloud DB foundation for Agent9 — needs schema extension, not replacement. |
| Neon PostgreSQL | LIVE | Secondary backup only. Could be repurposed or dropped. |
| Component | Status | Notes |
|---|---|---|
| Cloudflare Pages | WIRED (per git status context) | Static site hosting, auto-deploy on push. Suitable for Agent9 marketing site and seller onboarding flows. |
| Cloudflare Worker | BUILT, NOT DEPLOYED | Telegram + state fallback when Mac is down. |
| Mac local | PRIMARY | All critical services run on one Mac. This is a single point of failure for a product. Cloud migration is required before scale. |
| GitHub Pages | LIVE for AiNFLGM | Static hosting precedent exists. |
| Component | Status |
|---|---|
| Stripe credentials | IN LOCKER — not connected to any Agent9 product |
| Stripe Identity | NOT INTEGRATED — but this is the recommended KYC path for buyer ID verification |
| Any seller authentication | NONE |
| Any buyer authentication | NONE |
Seller Onboarding Seller creates an account, enters property details, uploads photos. Compensation agreement signed digitally. This is a web form + Stripe payment capture + e-signature. Buildable in 3-4 weeks. DocuSign API or Dropbox Sign API are both ~$25/mo for low volume.
MLS Submission This is the hardest single dependency and it is NOT buildable — it requires a licensed Ohio real estate broker to co-list. Ohio Division of Real Estate requires MLS listings be submitted by a licensed broker. Agent9 is not a licensed broker. A broker-of-record partnership is the only legal path. Options: (a) Ohio Broker Direct, Buckeye Flat Fee MLS, or a similar flat-fee broker partnership where Agent9 pays $50-150 per listing for MLS access; (b) or the Owner pursues a licensed Ohio real estate broker relationship. This is a PARTNER DEPENDENCY that cannot be built around. Without this, Agent9 cannot place listings on Zillow, Realtor.com, or the Ohio MLS. Timeline to partner: 2-6 weeks to find and execute an agreement — IF the Owner starts outreach immediately.
Buyer-Facing Search / Listing Display Two options: (a) rely entirely on Zillow/MLS syndication (Agent9 listings appear on Zillow via the broker partner's MLS submission — no build required); (b) build an agent9.com listings page. For Q3 MVP, option (a) is the right call. Build the listings page in Year 2 after MLS data volume justifies it.
AI Buyer Communications — Inbound Voice The existing voice-server.mjs + ElevenLabs + Twilio stack IS the foundation here. What needs to be built: a new Twilio number per listing, provisioned at listing creation; a new ElevenLabs voice persona for Agent9 (separate from Jasson's personal AI); a prompt system that takes property details as context and handles buyer qualification calls. Estimated build: 3-4 weeks.
AI Buyer Communications — Inbound Text / Email The comms-hub email and SMS receive paths exist. What needs to be built: multi-tenant routing (each listing has its own phone/email, not Jasson's personal channels), CRM-style lead tracking by listing, automated response templates, AI escalation logic. Estimated build: 4-5 weeks.
Showing Scheduling + Lockbox Integration
Buyer requests a showing. System checks seller-defined availability, books the slot, issues a time-limited lockbox code. Lockbox vendor: Igloohome (published developer API, PropTech partnerships confirmed, $100-200/device hardware). Stripe Identity for buyer ID verification before code is issued ($1.50 per verification). Estimated build: 4-6 weeks. Vendor selection and hardware sourcing must start now — lead time on Igloohome developer access and device procurement is unknown.
Offer and Document Generation AI drafts Ohio-compliant purchase agreements from stated terms. This requires: a vetted Ohio purchase agreement template (requires attorney review, one-time ~$500-1,000), an AI fill-in layer (Anthropic API with structured prompting against the template), and an e-sign delivery step. Estimated build: 3-4 weeks after template is reviewed.
County Auditor Deed Sweep (Enforcement) Automated monthly sweep of Ohio county auditor records to catch off-platform closings. This is a data engineering task — 88 counties, inconsistent formats. For Q3, wizard-of-oz this entirely: manually check deed records for the first 50 transactions using a paralegal or Jasson. Build the automation in Q4 once transaction volume justifies the engineering cost.
Stripe Payment Integration $999 fee collected at first showing confirmation. Stripe checkout session, webhook to mark the listing as paid. 2 weeks to wire. Stripe credentials are already in the Locker — this is the most-ready gap on the list.
Compliance: Ohio Real Estate Commission Agent9 is not a licensed brokerage and the business model does not require one if structured carefully (flat fee for communications services, not for brokerage representation). However, an Ohio real estate attorney must review the business model and confirm Agent9 is not inadvertently practicing real estate brokerage under ORC 4735. One-time cost: $800-1,500. This is a non-negotiable gate before first transaction.
| Gap | Build vs. Buy/Partner | Rough Cost | Timeline | Q3 Critical Path? | Wizard-of-Oz Option |
|---|---|---|---|---|---|
| Ohio MLS broker partnership | Partner only — cannot build | $50-150/listing submission fee | 2-6 weeks to secure | YES — blocking | None. Cannot list without this. |
| Seller web onboarding (form + Stripe + e-sign) | Build | $500 dev + $25/mo DocuSign | 3-4 weeks | YES | Google Form + manual Stripe invoice works for first 10 listings |
| AI inbound voice (Twilio + ElevenLabs, multi-tenant) | Build (existing stack) | $200-300 dev | 3-4 weeks | YES | Forward calls to a personal number temporarily |
| AI inbound text/email (multi-tenant routing) | Build (existing stack) | $300-400 dev | 4-5 weeks | YES | Manual response for first 10 listings |
| Showing scheduling + lockbox (Igloohome) | Build + hardware | $200 dev + $100-200/lockbox device | 4-6 weeks | YES | Seller controls their own lockbox temporarily |
| Buyer ID verification (Stripe Identity) | Buy | $1.50/verification | 1-2 weeks to integrate | YES | Manual ID photo collection by text |
| E-signature integration | Buy | $25-50/mo | 1-2 weeks | YES | DocuSign free tier for first 3 transactions |
| Ohio-compliant purchase agreement template | Buy (attorney review) | $500-1,000 one-time | 2 weeks | YES | Use Ohio Bar Association standard form temporarily |
| Stripe payment connect (fee collection) | Build | $200 dev | 2 weeks | YES | Manual invoice acceptable for first pilot |
| Attorney review (ORC 4735 biz model) | Buy | $800-1,500 | 2-4 weeks | YES — blocking gate | None. Do not transact without this. |
| Option C showing-agent network | Partner + build | Significant (recruit, vet, 1099, Stripe Connect) | 8-12 weeks minimum | NO — Year 2 | Offer Option A (seller-led) as the only alternative to Option B at launch |
| Lockbox dropship-with-markup pipeline | Build + vendor contract | Unknown — needs vendor negotiation | 6-10 weeks | NO — revenue layer, not core to MVP | Sell lockboxes at cost or direct buyer to Amazon at launch |
| Dedicated showing-agent buyer subscription | Build | Significant | 10+ weeks | NO — Year 2 | Not offered at launch |
| Mobile app (native iOS/Android) | Build | $40,000-80,000+ (native) | 4-6 months minimum | NO | PWA on launch — see Section 5 |
| County auditor deed sweep (automated) | Build | $1,000-2,000 dev | 6-8 weeks | NO — manual WoZ acceptable | Manual check, paralegal-assisted, for first 50 transactions |
| Escrow integration (Earnnest / Qualia) | Partner | $0-500 setup | 2-4 weeks | NO — title partner handles this | Route buyer to standard title company directly |
| County deed recording | Partner (title company) | Included in title fees | N/A — title partner does this | Not built by Agent9 | Refer to Ohio title company partner |
| Seller subscription / buyer subscription billing | Build | $300-500 dev | 3-4 weeks | NO | Not at launch |
| MLS Zillow syndication (auto-push) | Via broker partner | Included in broker partnership | Dependent on partner | YES (through broker) | Manual listing creation on Zillow FSBO directly |
22 weeks from now is Q3 2026. There is no world where a 2-person team ships a production-grade multi-tenant real estate platform with MLS submission, lockbox hardware, ID verification, e-signature, offer generation, and Stripe Connect in 22 weeks while also running a marketing operation. It is not a question of effort — it is arithmetic.
What IS achievable by Q3 2026 with a 2-person team:
A live, bookable, real-transaction-capable demo platform that can handle the end-to-end flow for 10-20 sellers, with wizard-of-oz on the back-office complexity. Specifically:
What is wizard-of-oz at launch and acceptable:
What CANNOT be wizard-of-oz (hard blockers if absent):
Three partners are required before the first transaction can close:
These are not buildable. If any of them are delayed more than 6 weeks, Q3 demo is at risk. The Owner should be making warm contact with candidates for all three this week.
Native iOS + Android: do not build for Q3 2026.
Native app development for both platforms with an unfamiliar codebase and a 2-person team is a 4-6 month minimum timeline, $40,000-80,000 minimum cost if contracted, and introduces App Store approval friction (Apple real estate apps require specific review criteria, and Apple's review times for new categories average 1-3 weeks). Building native apps before the web product is validated is a classic startup trap.
React Native or Flutter (cross-platform native): marginally better but still wrong for Q3.
Still requires 10-14 weeks minimum for a quality cross-platform implementation. The advantage over PWA is access to push notifications and camera for ID document upload — both valuable. But not by Q3.
PWA (Progressive Web App): the right call for Q3 2026.
A well-built PWA on a mobile-responsive web stack delivers:
AiNFLGM already runs as a PWA — the Vite + PWA plugin setup exists in this codebase. The precedent and toolchain are in place.
Verdict: PWA at launch, React Native or Flutter in Year 2 once the web product is validated and seller/buyer volume justifies the investment. Do not let "no native app" be positioned as a weakness — Opendoor ran their self-tour model on mobile web before building native apps. The showing flow is functional on mobile Safari/Chrome without a native wrapper.
Yes — with two mandatory gates cleared first.
The Owner asked: is the next natural step after the marketing plan to develop the first operational demo of the website?
The answer is yes, but not in a vacuum. The demo website is meaningless — and potentially harmful — if it points to a product that cannot deliver on its promise. The two gates that must be cleared before demo build starts:
Gate 1: Ohio attorney review. The business model needs a clean legal opinion that Agent9 is not inadvertently operating as an unlicensed broker under ORC 4735. This costs $800-1,500 and takes 2-4 weeks. If the attorney finds a structural issue, the product design changes. Build nothing until this is in hand.
Gate 2: Ohio broker-of-record contact initiated. The demo is hollow without a clear answer to "how do we get listings on the MLS?" This is a partner relationship, not a build task. The Owner needs to make this contact immediately — it has the longest lead time of any gap on this list.
If those two gates are in motion, the demo MVP is:
Minimum compute / hosting to stand this up:
No new infrastructure category is required. Every piece is either live or accessible from the Locker. The build is wiring, not invention.
This is the honest recommended path from today to a working demo that can book real Ohio transactions.
Weeks 1-2: Legal and Partner Foundations
Weeks 2-4: Core Platform Build
Weeks 4-6: Communication Layer and Showing Flow
Weeks 6-8: Offer Flow and Demo Hardening
At Week 8 (early Q3 2026): A seller in Columbus can sign up at agent9.com, pay $999, have their listing submitted to Ohio MLS by the broker partner, receive AI-answered buyer calls and texts, schedule a verified self-tour, and receive an AI-drafted offer through the platform. A human (Jasson) reviews AI output before any document is sent. County auditor enforcement sweep is manual. Option C showing agents are not available. Lockbox dropship is not available. But the core product works, real money flows, and the first 10-15 transactions can be completed.
That is Q3 2026 launch. It is honest. It is achievable. It is not the full vision — but it is enough to prove the model and earn the right to build the rest.
| Assessment | Confidence |
|---|---|
| Existing 9enterprises infrastructure (comms, voice, Supabase) | HIGH — read directly from dependency-map.json and codebase |
| Gap list and build estimates | MEDIUM — estimates based on comparable integrations; actual complexity depends on Igloohome API quality and attorney findings |
| Ohio broker partnership feasibility | MEDIUM — Ohio flat-fee brokers exist and the arrangement is common, but negotiation timeline is variable |
| Q3 demo MVP achievability (wizard-of-oz version) | HIGH — this is a 2-person team working a focused sprint on a bounded scope |
| Q3 full-scale production platform | LOW — not achievable with 2 people in 22 weeks, full stop |
| PWA vs. native app verdict | HIGH — well-established best practice for pre-product-market-fit mobile |
Prepared by UNO — Research Team Lead | April 25, 2026 Source files: CLAUDE.md (architecture), agent9/marketing/marketing-plan.md, agent9/vision/agent9-walkthrough-script.md, agent9/research/competitive-sweep.md, docs/dependency-map.json