The marketing system your CMO runs.
Alex installs it.
Outreach, paid control, retention, SEO content, analytics and dashboards, executed by agents your own marketing lead approves. Two weeks to install and train. Then it runs.
Needs your decision
What runs underneath
Each line replaces a person a growing company would hire. They share one client config, one CRM and one approval queue, so nothing gets re-explained between specialists.
- OutreachKOL, sponsor, partner and listing sequences over email and WhatsApp Cloud API. Reply watcher stops follow-ups and drafts an answer.live
- Paid controlNetwork monitors, fraud detector, rules that pause, resume and move budget with a human yes.live
- RetentionBehavioral drips on Mailgun and Mailjet, warm-up, SPF/DKIM/DMARC before every send, language by geography.live
- SEO contentKeywords from live search demand, articles with FAQ schema in five languages, published on schedule.live
- AnalyticsSurvival-weighted LTV, marketing profit, CAC by source, anomaly alerts.live
- LaunchesA 53-task framework per product, feature or region.live
- AuditSpend → traffic → signups → payers → LTV → profit, ten checks, scored.live
- NewsroomMulti-channel Telegram publishing with a house voice and local-time cadence.live
- DashboardsBoards provisioned from templates and your config, with backfill guards.in rollout
Built inside companies before it was a product
Numbers are ratios and timings. Client data stays with the clients.
PrimeXBT
Hours per processed lead, after KOL and introducing-broker sourcing moved into one CRM pipeline. $7M quarterly budget run at up to 120% ROI, a 30+ person team.
Tarija Tech Week
From zero to 200+ sponsor leads in sequences. 81 WhatsApp touches and 159 emails on day one; two sponsorship deals closed over WhatsApp days before the event, Binance as main sponsor.
Telegram AI aggregator
Three outreach machines run weekly with one person reviewing replies. Attribution moved from cost per action to revenue; the cheapest campaign turned out to be the worst money.
How it starts
Not a chatbot that writes posts. A control loop with your CMO at the decisions that cost money, installed by the person who built it.
Connect
CRM, analytics, ad accounts, email domain, site. Client config written: brand, voice, caps, rules of engagement.
Switch on
The modules that matter for your stage go live. Deliverability is fixed before any email leaves.
Train your CMO
Hands-on, on your real data: reading alerts, approving spend, steering outreach and content.
Run
Your CMO runs it from the queue you saw above. Alex stays on call in a Telegram thread for configuration and new modules.
Pricing
One subscription covers installation, your CMO's training, the agents and every module. Ad spend and tools you already own are yours.
- Audit of current marketing, ten checks
- Client config and connections
- Your CMO trained on your data, two weeks
- First modules live within two weeks
- Every module, no per-module fees
- Infrastructure and agents' compute
- Updates and new modules as they ship
- Alex on call for your CMO
- Cancel any month, configs stay yours
No CMO yet? Alex runs the system for the first 90 days while you hire, then trains the person you bring in.
Alex Khvoinitskii
Twenty years in marketing, dozens of companies from the inside, hundreds consulted. VP Growth & Partnerships at PrimeXBT, CMO of a Telegram AI product with 30M+ users and of an adult AI companion app, organizer of Bolivia Blockchain Conference and Tarija Tech Week.
Since 2022 I have been turning marketing work into systems, first content, then outreach, then analytics, until whole functions ran on agents with a human at the money decisions. AI CMO OS is those systems, cleaned of client data, installed for your CMO. I am the architect, not your operator.
Talk to Alex
Tell me what you sell, where customers come from today, who runs marketing on your side and what you expect in 90 days. A setup plan comes back within a day.