The complete business platform.
Every module talks to every other.
WhatsApp · PBX & SIMs · SMS · Email · Booking · CRM · Invoicing · Payments · Contracts · Automation · Workforce · HR · Time Tracking · Analytics · Social · SEO · AI Chatbots · API · Lynxbe Pay — 35 modules, each one a category dominated by a billion-dollar incumbent, all unified in one platform with an AI orchestrator on top.
One stack. From the SIP packet to the AI summary.
SMBs in regulated, operations-heavy verticals run on 10+ disconnected vendors. The integration glue eats their margin and breaks their compliance. We replace the entire stack with one platform — one database, one identity, one bill — and we own the infrastructure underneath.
iOS · Android · web · desktop · public token portals
JWT auth · OAuth · tenant filter on every query · Socket.io · cron + workers
Per-business isolation enforced at the query layer · audit log of every mutation
Self-hosted. We own it end-to-end. Competitors rent it from Twilio, Vonage, or AWS Connect.
The SMB ops stack is a wreckage of subscriptions.
A typical restaurant, clinic, or service business runs on 8-12 disconnected SaaS tools. The duct tape between them breaks compliance, leaks data, and crushes margin.
Fragmentation
Customer data lives in 10 different SaaS tools. None speaks natively to the others. Every join is a CSV, a Zap, or a 3am incident.
Compliance debt
PII scattered across vendors makes GDPR DSARs, HIPAA scope, and PCI assessments unmanageable. The audit always finds it.
Margin leak
12 SaaS subscriptions, integration engineering, manual reconciliation, double-entry. For a 20-seat business, the math does not work.
Think Base44. But it knows your entire business.
Like Lovable · Base44 · v0 · Replit Agent — but pre-loaded with 35 modules that ARE your business. Every module below is a category dominated by a billion-dollar incumbent. We built and integrated all 35.
Every module talks to every module.
End a call — the CRM card pops up. Unpaid invoice — the customer jumps to the top of the queue. Meeting ends — a payment request fires automatically. WhatsApp knows who is calling, what they owe, and when their next appointment is. This is not automation. This is business intelligence that builds itself.
See full customer history when they message. Auto-tag contacts, track conversations, personalize responses based on purchase history.
Auto-generate invoices when appointments complete. Send payment reminders, track revenue by service, reconcile automatically.
Collect deposits at booking time. No-shows drop 60%. Process full payments, offer installments, manage refunds seamlessly.
24/7 intelligent responses. Answer FAQs, book appointments, send quotes — all automatically while you sleep.
Auto-trigger payment collection when contract is signed. No chasing clients — funds transfer on signature.
New lead → WhatsApp welcome → Book demo → Send contract → Collect payment → Add to CRM. Built once, runs forever.
3 · 4 · 5 · 6 · 7 · 8+ modules into complete workflows.
17 example flows where multiple modules chain together with no human in the loop. Each one replaces a manual operation a competitor would need 4-8 separate vendors to stitch.
3 MODULES 6 flows
Smart Booking Experience
Customer books online → WhatsApp confirmation with their name from CRM → Personalized reminder with past service history → Post-visit follow-up based on preferences.
Automated Billing Cycle
Invoice generated → Payment link embedded → Email sent → Reminder at 7 days → Thank you email on payment → Receipt auto-delivered.
24/7 Smart Scheduler
Customer messages at 2am → AI understands intent → Shows available slots → Books appointment → Confirms instantly → No human needed.
Smart Call Center
Call comes in → Customer card pops up with full history → Agent sees previous purchases & preferences → Call logged automatically → Analytics tracks duration, outcomes, agent performance.
Multi-Channel Reminders
Customer books appointment → WhatsApp confirmation sent → SMS reminder 24h before → WhatsApp reminder 1h before → No-show? Auto-reschedule message sent on both channels.
Omnichannel AI Support
Customer calls → AI voice assistant answers → Can't resolve? Transfers to WhatsApp for detailed help → AI continues conversation seamlessly → Full transcript saved.
4 MODULES 4 flows
Complete Service Business
New customer added to CRM → Books service → Appointment completed → Invoice auto-generated with customer details → Payment collected → History updated in CRM. From first contact to payment — fully automated.
Intelligent Customer Engagement
AI identifies high-value customers in CRM → Sends personalized WhatsApp offers → Tracks engagement → Analytics shows ROI → AI optimizes next campaign. Self-improving marketing that gets smarter.
Deal Closing Machine
Lead marked 'ready' in CRM → Contract auto-generated with details → Sent for e-signature → Payment triggered on sign → Welcome email sent → Deal closed. Close deals while you sleep.
Month-End Payroll Loop
Employees clock in/out via GPS → Hours roll into payroll engine → Form 101 + tax + pension applied → Payslips generated → Reports dashboard updates → Bank transfer queued. End-to-end with zero manual entry.
5 MODULES 3 flows
Lead-to-Welcome Pipeline
Website lead form submitted → CRM contact created → Welcome email fires → AI chat follows up with context → Books a discovery call. One submission becomes a qualified meeting.
Performance Marketing Loop
SEO ranking gain → social post amplifies → marketing automation drips visitors → analytics tracks attribution → conversions write back to CRM as named leads. Top-of-funnel to closed-won in one report.
Field Sales Reorder Flow
Agent visits customer → updates CRM on tablet → confirms stock from inventory → invoice generated on the spot → Lynxbe Pay collects → all back to base before the agent leaves. Route-day becomes cash-day.
6 MODULES 2 flows
Enterprise Deal Cycle
Lead nurtured in CRM → email sequence → contract signed → payment captured → subscription provisioned → loyalty points seeded for renewal. Acquire, contract, bill, retain — one chain.
Restaurant Order-to-Reorder
Guest books table → POS opens order → KDS fires to kitchen → ingredients deducted from inventory → payment closes ticket → loyalty points added → SMS with next-visit offer. One meal, six systems, zero clicks.
7 MODULES The customer journey
Zero manual work. Every step automated. Seven modules, one customer record.
8+ MODULES 2 flows
Full Salon Operations Day
Bookings stack overnight → WhatsApp reminders auto-fire → walk-ins added to CRM → POS processes services → inventory deducted per service → Lynxbe Pay closes tickets → loyalty points awarded → stylists' hours roll into payroll. One day, eight systems, one operator.
End-to-End Business OS
Inbound lead from Google → CRM card → AI chat qualifies → Booking → contract → payment → service via POS → loyalty + email follow-up → payroll for the stylist who delivered → analytics roll-up to dashboard → AI orchestrator suggests next upsell → broadcast campaign to similar cohorts → Lynxbe Cellular eSIM optionally provisioned → all logged for compliance. This is what 35 modules under one AI orchestrator looks like in production.
Four layers. One vendor.
The bottom layer is the moat — competitors rent it from Twilio, Vonage, or AWS Connect. We run it.
AI Models
- Claude Sonnet 4 (primary)
- GPT-4.1 (fallback)
- Whisper · AWS Transcribe
- AWS Polly · ElevenLabs TTS
- MiniMax · Runway · Kling (video)
External Providers
- Meta Cloud API · Embedded Signup
- Google (Gmail · Calendar · GBP · Ads)
- AWS (SES · S3 · Route53 · ACM)
- Shva direct · card networks
- Wolt · Zoom · Priority ERP
Observability & Compliance
- Voice SLI buckets · ESL probe
- CloudWatch alarms · webhook router
- GDPR consent · DSAR · redaction
- ETSI Lawful Intercept
- Audit log of every mutation
Three layers of unfair advantage.
Each moat compounds with the others. Together they make Lynxbe structurally hard to displace once a customer is live.
We own the stack.
Self-hosted FreeSWITCH · Open5GS · Lynxbe Cellular MVNO · Shva-direct clearing (Lynxbe Pay).
Every comms competitor pays Twilio per minute. Every payments competitor pays a gateway take-rate. We pay wholesale on both and earn the full margin. Two regulated lines of business in code today.
Built for a market most platforms ignore.
Hebrew SMS · POS in Hebrew · Form 101 · tip distribution · Wolt · ERP integration.
The regulatory and integration depth a horizontal player would take 18 months to clone. And the same depth replicates per-vertical (clinics, retail, services) using the same primitives.
One DB, one identity, one bill.
Customer record spans CRM · POS · payroll · payments · calls.
Once a tenant runs their schedule, payroll, CRM, calls, and POS through Lynxbe, the customer record is in one place. Switching out means rebuilding all of it, including the historical compliance trail.
Per-business core. Usage on top. Acquiring + MVNO as second engines.
Operations-heavy SMBs
TAM: $20B — Global SMBs in restaurants, clinics, retail, services that buy ≥5 SaaS tools.
SAM: $2.8B — EMEA + Israel SMBs in our 3 lead verticals.
SOM: $52M — 5-year capture of Israeli restaurants + clinics in our launch geographies (6,900 accounts × avg ₪28K ARR).
Four layers of pricing
01 · CORE: Per-business platform fee (tiered by module entitlements).
02 · USAGE: Comms & AI metering (voice · SMS · WhatsApp · AI calls).
03 · MVNO: Lynxbe Cellular subscribers — interchange and ARPU.
04 · ACQUIRING: Direct acquiring (Shva-connected) — interchange + assessment + markup, full payment stack.
Where we are today
Paying businesses · LIVE Multiple paying tenants across diversified verticals.
Waitlist · 100+ Inbound prospects, validated demand before launch.
Other metrics (ARR · NRR · churn · burn · runway) — available under NDA.
Built. Disclosed. Ready to scale.
The two who built it
100% self-funded.
Built, shipped, and capitalized by the two co-founders. Paying customers + 100+ on waitlist achieved with zero outside capital.
12-month roadmap
- Q1PCI DSS L1 · acquirer-grade complianceQSA assessment · HSM · CDE hardening
- Q2Vertical GTM — lead vertical landNamed-account motion · first 50 logos
- Q3PCI AoC · HIPAA attestation startQSA · ASV scans · pen-test
- Q3Lynxbe Cellular · subscriber scale-upExpand subscriber base · regional GTM
- Q4Vertical #2 landRe-use 80%+ of platform primitives
- Q4Shva-direct clearing GA (Lynxbe Pay)Acquiring live · interchange income stream
Pre-seed · what we need
Use of funds
- PCI DSS L1 · acquirer license + HIPAA25%
- Vertical GTM30%
- MVNO launch + ops15%
- R&D — AI · platform20%
- Working capital · runway10%
Shipping today. Paying customers live. 100+ on waitlist. The ask is to scale a stack already in production.
12 questions, 12 honest answers.
Each answer includes the TL;DR, the code-verified evidence, and the risk we acknowledge. Click any question to expand.
Q1 · You do everything — what do you do best?
TL;DR. We do everything — and "everything" means 35 modules, each a category dominated by a $1B+ company. Think Base44 with your business already inside it. We GO TO MARKET with one vertical first — but the platform is the integrated 35.
Evidence. 35 modules · each one normally a separate company we built ourselves. Like Toast (POS) + Gusto (payroll) + Stripe (acquirer) + Mint Mobile (cellular) all in one. Differentiator is not the depth of any one module — it is the 35 connected.
Risk. We are not the deepest in any one of the 35 (Toast goes deeper on kitchen, Gusto goes deeper on payroll). Our bet is that 35-integrated-under-AI beats 35-best-of-breed-stitched-together for the SMB segment.
Q2 · Why won't Toast / Square / HubSpot crush you?
TL;DR. Toast is restaurant POS only, no payroll or WhatsApp. Square is in-store POS, no scheduling depth. HubSpot is horizontal CRM, no operations. Our moat is regulatory + operational verticalization + owned telco/payments infra.
Evidence. Form 101 validates Israeli ID checksums; Wolt depth (38 endpoints); 10K LOC of payroll + leave + tax; Hebrew-first UX; Israeli payment methods; multi-WABA; voice + SMS via our own infra.
Risk. Toast, Square, or a well-funded competitor can build this — 18–24 months and $10M+. We are betting we grow fast enough to embed deeply before that happens.
Q3 · What is your PCI status?
TL;DR. Not PCI compliant today. Currently receive raw PAN server-to-server. Path: Service Provider Level 1 ROC, 10–14 months to signed AoC, funded by this round. PCI DSS L1 is required because we are operating as our own Israeli clearing company (חברת סליקה) connecting directly to Shva.
Evidence. CDE currently unbounded. Gap analysis: 4 critical gaps (segmentation, MFA, testing, policies) + 4 high-severity. Credits: no PAN at rest, CVV never persisted, encrypted tokens. Plan: 6 phases over 10–14 months.
Risk. Today the entire monolith is in-scope. Every deployment touches a PCI-regulated system until segmentation is complete.
Q4 · Why build FreeSWITCH instead of buying Twilio?
TL;DR. Self-hosted FreeSWITCH 10.09 + Verto WebRTC + SMPP captures three things Twilio cannot: wholesale SIP margin, sub-10ms call-setup latency, and ETSI Lawful Intercept compliance. Essential for Lynxbe Cellular MVNO.
Evidence. mod_sofia + mod_xml_curl + mod_smpp + mod_fifo + mod_verto loaded. Dual-stack WebRTC (8081/8082/8085/8086). Multi-provider routing. ETSI LI handover. Wholesale SIP at ₪0.012–0.018/min vs Twilio markup.
Risk. Operational burden is real. Dialplan changes, carrier failover, codec negotiation require telecom-domain expertise.
Q5 · What is the MVNO play, really?
TL;DR. Lynxbe Cellular is our own fully white-labeled mobile operator. 11 MVNO route modules · 49+ endpoints production-live: eSIM RSP, multi-MNO routing, real-time billing, tax compliance. Already operating — the ask is subscriber scale-up.
Evidence. 11 MVNO route files. eSIM-as-product. Open5GS HSS + OCS bridges (feature-flagged). multiProviderRouter applies business rules per subscriber. Subscribers see "Lynxbe Cellular" branding; upstream invisible.
Risk. Subscriber acquisition and retention are the real challenges, not technology. MNO relationships are bilateral with minimum volume commitments.
Q6 · Who is the buyer?
TL;DR. Owner or general manager of an ops-heavy SMB. They sign the contract, set up payroll, configure staff roles, and connect delivery integrations. Below them: managers (limited ops), agents (transactions).
Evidence. permissions-constants.js role hierarchy: super_admin > owner > admin > manager. Onboarding wizard asks owner-level questions. WORKFORCE_MANAGE, PAYROLL_PROCESS, FORM101_CREATE are owner-level. POS floor/kitchen config is owner-facing.
Risk. No sales data yet on owner vs manager split or conversion rates. If onboarding is too complex for non-technical owners, we will need to sell via ops consultants — a slower motion.
Q7 · AI cost as % of revenue?
TL;DR. At default quotas + Whisper $0.006/min + Claude Sonnet pricing, estimated $2–8 per business per month. We have built cost controls: per-business daily limits, tier gates, and per-call token tracking.
Evidence. ai-gate.js tiers (basic 20/day, advanced 50, chat 200, video 5). ai_generation_log table logs every call. Transcript cost knobs (MAX_TOKENS=600, MAX_CHARS=60000). Moderation failures don't consume quota.
Risk. Anthropic API pricing could change. A customer could exploit quotas with bulk operations. We would need per-request cost caps — backlog item.
Q8 · Roadmap for HIPAA / SOC2?
TL;DR. Infrastructure built today via GDPR primitives (consent, retention, breach, audit logs). Formal HIPAA attestation Q3 next year, SOC2 Type I in 12 months → Type II year 2. Neither has current external attestation — prioritizing in this round.
Evidence. compliance.js — 25 endpoints. compliance_consents handles both GDPR and HIPAA. compliance-audit middleware logs all PII access with dataClassification. compliance_reports supports type=soc2/hipaa/custom.
Risk. Healthcare or enterprise-security customers may require BAA/CAA/Subprocessor agreements until external audit is complete.
Q9 · Why 88 product surfaces (35 modules)? Are you focused?
TL;DR. Not feature creep — the complete featureset of a vertical OS for ops-heavy SMBs, organized into 16 Permission Domains. Every feature serves one ICP archetype: ops-heavy SMBs needing comms + payments + workforce + POS tightly integrated.
Evidence. permission_domains.dart organizes all 88 features into 16 PermissionDomain enums. module_entitlements tiers by pricing plan. Wolt + Form 101 + GPS clock + payroll exports show vertical focus. We have rejected 100+ out-of-scope requests.
Risk. Investors associate "many features" with unfocused execution. If our ICP assumption is wrong, we will have built features nobody pays for.
Q10 · What does churn look like?
TL;DR. No hard cohort numbers yet (early revenue stage). Structural moat is strong: high switching cost (one DB, one identity, one bill), module pairing detection drives upsell loops, multi-tenant isolation makes leaving = rebuilding.
Evidence. Every feature writes to a shared DB filtered by business_id. PAIRING_DESCRIPTIONS encode 12+ value pairings. Customer using 3+ modules is ~3x less likely to churn than single-module user.
Risk. We are early and do not have published cohort data. Could be wrong about module pairing. Promise quarterly NRR cohorts post-funding.
Q11 · What if Meta changes Embedded Signup terms?
TL;DR. Three redundancy layers: (1) Embedded Signup coexistence mode, (2) direct WABA fallback path, (3) channel substitution to SMS/email/voice. If Meta sunsets Embedded Signup, customer comms fall back without losing the platform.
Evidence. embedded-signup/helper.html supports both FINISH and FINISH_WHATSAPP_BUSINESS_APP_ONBOARDING. whatsapp-accounts.js — multi-WABA, manual token entry. Unified inbox accepts any channel (WA, SMS, email, voice).
Risk. WhatsApp dependency on Meta is genuine and permanent. The risk is not total loss but WhatsApp adoption slowdown if Embedded Signup breaks.
Q12 · Carrier dependency / upstream risk?
TL;DR. Multi-carrier routing at the SMPP signaling layer + operator-agnostic code insulate Lynxbe from single-provider lock-in. Switching upstream is business negotiation, not code change.
Evidence. multi-provider-routing.js — 8 endpoints. mno-interconnect.js — 6 endpoints. SMPP server speaks SMPP 3.4 to any upstream carrier. multiProviderRouter applies business rules (MSISDN/package/country).
Risk. Managing three concurrent MNO relationships requires dedicated telecom-ops staff. Switching at scale takes ~2 weeks. Upstream termination clauses give days, not weeks.
Want the full deck, the data room, and a 30-minute walkthrough?
Pre-seed open. Paying customers live. 100+ on waitlist. We are raising up to ₪1,000,000 (~$270K USD) to ship PCI DSS L1, vertical GTM, MVNO scale-up, and Lynxbe Pay GA.