New MailCore product · in development

MTA-Hub — an AI control panel for PowerMTA, MailerQ and your whole SMTP infrastructure

Design, connect, warm up and monitor MTA clusters from a single panel. MTA-Hub brings together servers, IPs, domains, vMTAs, queues, configurations and Prometheus metrics — while AI audit tells you what to do next.

🚧 The product is in active development — early subscribers get a discount when sales open
mta-hub · overviewPRODUCTION
23MTA Nodes+3 this week
22Online95.8% uptime
128KQueue−4.2%
92AI Scorehealthy
server-01· PowerMTA · prod-euHEALTHY
server-03· PowerMTA · warmup-aWARMUP
server-11· MailerQ · prod-usRISK
AI Metrics Auditor: on server-11 the 4xx growth coincided with a rise in outbound connections — I recommend slowing down the prod-us pool and reviewing the latest config changes.
PowerMTA + MailerQ + SMTP in one registry Prometheus monitoring out of the box Configurations with diff and rollback AI audit and recommendations
The problem

Infrastructure grows — and manual management turns into chaos

One or two SMTP servers can still live "in your head". But once PowerMTA and MailerQ nodes grow into dozens, along with hundreds of IPs and domains, — without a central panel you lose both visibility and control.

01

Scattered servers

IPs, domains and vMTAs are spread across panels, spreadsheets and admins' heads. There is no single registry — nobody knows exactly what runs where.

02

Configs drift apart

PMTA and MailerQ settings are duplicated by hand and inevitably drift between nodes. Which server runs which version is anyone's guess.

03

Queues swell unnoticed

While you watch one server, a queue grows on another. The problem surfaces after the fact — when delivery has already dropped.

04

Failure source unclear

4xx/5xx pour in, and without consolidated metrics and logs you can't tell: is it the config, IP reputation, content, or node overload.

05

Edits straight in files

Changes go over SSH into live configs — no validation, no review. One typo costs hours of downtime.

06

No rollback, no analysis

No version history, no audit log, no AI hints. Every incident is investigated from scratch and from memory.

Important to understand: Protosender, Mumara and Interspire are great at sending campaigns — but deep infrastructure management underneath them is not their job. That takes a dedicated command center.

The solution

MTA-Hub turns scattered servers into a managed system

One pipeline: from the goal and an AI plan — to monitoring, audit and optimization. All in one panel, with history and the right to make mistakes.

Without MTA-Hub

  • Configs edited by hand over SSH — each server has its own
  • Queues and errors live in different windows and panels
  • Warm-up by eye, with no risk control
  • A dangerous change can't be rolled back — versions aren't stored
  • Incidents get caught after the fact, when delivery has already dropped

With MTA-Hub

  • Configs via templates, diff and validation — one standard across all nodes
  • Queues, errors and metrics — in one panel across all servers
  • Gradual warmup with segments and risk monitoring
  • One-click rollback + a full audit log of changes
  • AI Core signals early — before the incident becomes visible
The full cycle — in one pipeline
01Goal 02AI infrastructure plan 03Servers 04PMTA / MailerQ 05Configurations 06Warmup 07Production 08Monitoring 09AI audit 10Optimization

From goal to optimization — AI Core plans the infrastructure on the way in and audits how it runs on the way out.

Core Platform

Eight modules — full infrastructure control

From an AI plan and a server registry to safe configurations, warm-up and exporting ready pools into your sending platform.

01 · Planning

AI Infrastructure Planner

You pick the goal, volumes, list source and MTA type — AI Core generates an infrastructure plan and a diagram.

  • Goal and expected volumes
  • Sending model and MTA type
  • Ready plan + diagram
02 · Inventory

Server & MTA Inventory

A central registry of the whole infrastructure — nothing lives "in spreadsheets and heads".

  • Servers, IPs, domains, vMTAs
  • MTA type, provider, location
  • Role, status, warmup/production
03 · Observability

Prometheus Monitoring

PowerMTA metrics via the native exporter — consolidated across all nodes at once.

  • Traffic and queues
  • SMTP connections and errors
  • Latency and first-attempt
04 · Control

Queue & Error Control

Queues and errors under constant watch — problem nodes are visible immediately.

  • Queue size and dynamics
  • 4xx / 5xx / transient
  • Delays and problem nodes
05 · Configs

Configuration Management

PMTA and MailerQ configurations — as code: with review, history and the right to roll back.

  • Templates and a directive library
  • Diff + validation before deploy
  • Rollback and audit log
06 · Warmup

Warmup Manager

Controlled rollout of infrastructure into production — without burning fresh IPs.

  • Active segments
  • Gradual sending and limits
  • Risk monitoring → production
07 · Export

API Export

Ready production pools go straight into your sending platform.

  • Export to Protosender
  • Mumara / Interspire
  • Strategy ↔ infrastructure
08 · Audit

AI Infrastructure Audit

AI Core regularly reviews the whole picture and tells you what to fix first.

  • Metrics, queues, errors
  • Configuration changes
  • Risk patterns + recommendations
Monitoring

Prometheus shows what's happening. MTA-Hub explains what to do next

PowerMTA exposes metrics in Prometheus format. MTA-Hub collects them from every node, joins them with logs, webhooks and campaign data — and turns it all into one observable picture with AI conclusions.

Server health · all nodes5 nodes · live
NodeMTAPoolQueue4xx5xxLatencyStatus
server-01PowerMTAprod-eu12.4K1.2%0.3%820msHEALTHY
server-03PowerMTAwarmup-a48.2K4.8%0.9%1.9sWARMUP
server-07MailerQprod-us9.1K2.1%0.4%640msHEALTHY
server-11PowerMTAprod-eu62.7K7.3%2.1%3.4sRISK
server-14SMTPprod-asia5.8K1.0%0.2%510msHEALTHY

AI Metrics Auditor: on server-11 the 4xx growth coincided with a rise in outbound connections. I recommend slowing down the prod-eu pool and reviewing the latest configuration changes.

queue rising · 45 min4xx +7.3%config drift detected
How much is being sent

traffic_recipients_total — traffic volume per node in real time.

How big the queue is

current_queue_size — queue size in recipients and bytes: node load at a glance.

Are errors growing

smtp_errors_total — 4xx/5xx signal delivery and reputation problems before the complaints arrive.

Is latency rising

request_duration_seconds — whether the MTA node is overloaded right now.

Is delivery delayed

first_attempt_seconds — how fast the first attempt goes out after inject.

Is the node healthy overall

The combined metrics roll up into an AI Score — an overall verdict on the node's state.

AI Core

Not a copywriter — an infrastructure analyst

AI Core in MTA-Hub doesn't write emails. It plans, warms up, reviews configs and investigates incidents — like a seasoned deliverability engineer who watches every metric at once and never sleeps.

AI Infrastructure Planner

Turns your goal and volumes into an infrastructure plan and diagram.

AI Warmup Strategist

Picks the pace and segments for gradual warm-up.

AI Config Advisor

Explains directives and risks, recommends safe values.

AI Incident Analyst

Correlates metrics and events during delivery incidents.

AI Metrics Auditor

Regularly cross-checks metrics and queues — catches anomalies.

AI Risk Scoring

Scores the risk of going to production and of config changes.

AI Documentation Assistant

Prepares documentation for the infrastructure and decisions.

AI Recommendations

Concrete, safe steps for every finding.

ai-core · live audit streamconnected
queue

The queue on server-03 has been growing for the last 45 minutes.

correlation

The 4xx growth coincided with a rise in outbound connections.

warmup

Moving to production is premature: the bounce rate is above the acceptable level.

action

I recommend slowing down the pool and reviewing the latest configuration changes.

config

The server-07 configuration differs from the production-pool-v2 template.

Configurations

PMTA and MailerQ configs — safe, like code

Templates, a directive library with explanations, AI recommendations, diff before apply, validation, safe deploy and rollback. The core principle: no dangerous change is ever applied blindly.

1
Pick a directive

From the library — each explained with its purpose and risks.

2
Assess the risks

AI shows what makes the directive dangerous and which parameters are critical.

3
Set the scope

Values apply at the server / vMTA / domain / pool level.

4
Review the diff

Before/after changes are fully visible before anything is applied.

5
Validate and deploy

Syntax check and safe apply, recorded in the audit log.

6
Rollback if needed

Revert to any past version — in one click.

server-07 · pmta.confbeforeafter
42<domain production-pool-v2>
43-  max-smtp-out  50
43+  max-smtp-out  25
44-  smtp-delays  60s/60s/90s/120s
44+  smtp-delays  90s/120s/180s/300s
45   dkim-signature (selector=default)
46-  queue-rate  100/s
46+  queue-rate  40/s
47</domain>
AI: safer for warmup ✓ validated v2.3 → v2.4
Warmup Manager

Infrastructure goes live gradually and under control

Active segments, controlled pace, monitoring of errors and reputation risks — and only then production mode. No "cranked it to full — burned it in a day".

Segment A

Active in the last month

Hot segments from your sending platform — for the early warm-up stages: fast feedback, minimal bounces, reputation grows from day one.

Segment B

Active 1–3 months

Warm segments for scaling the volume: we verify the reputation holds on bigger traffic before fully switching to production.

Pool lifecycle · production-pool-v2
✓ Planning ✓ Warmup Production Paused Risk Incident
warmup plan · production-pool-v2
mta-hub> warmup plan --pool production-pool-v2 --segment A,B
✓ day 1–3: 5K/hr · segment A · monitor 4xx < 3%
✓ day 4–7: 15K/hr · segment A+B · monitor bounce < 2%

Reputation-aware sending — honestly: MTA-Hub doesn't "guarantee the inbox" — nobody can. It does something else: keeps pace, errors, queues and risks under constant control so your reputation grows instead of burning out.

Integrations

The infrastructure layer between sending and MTA clusters

MTA-Hub sits at the center of the ecosystem: sending platforms above, MTA nodes below, observability, CRM and AI Core alongside. All data and commands flow through one hub.

Protosender SENDING · API CONNECTED Mumara SENDING · API CONNECTED Interspire SENDING · API CONNECTED Prometheus · Grafana OBSERVABILITY · SCRAPING KendaliCRM BUSINESS · ATTRIBUTION MTA-Hub CONTROL CENTER registry · configs · warmup · monitoring AI Core INTELLIGENCE · PLAN / AUDIT / RISKS PowerMTA MTA CORE · METRICS + CONFIG MailerQ MTA CORE · API CONTROL SMTP nodes MTA CORE · INVENTORY commands flow down to the MTA nodes · metrics and statuses rise up · AI Core analyzes everything in real time

← scroll the scheme horizontally →

PS
Protosendersending platform
  • SMTP/MTA pool export
  • Strategy ↔ infrastructure
  • Warmup segments
  • Revenue attribution via KendaliCRM
MM
Mumarasending platform
  • SMTP pool export
  • Status synchronization
  • Sending infrastructure mapping
IS
Interspiresending platform
  • SMTP export
  • Pool grouping
  • Sending node management
KC
KendaliCRMbusiness
  • Sales attribution
  • Customer journeys
  • Revenue reports
PMTA
PowerMTAMTA core
  • Prometheus metrics
  • Configuration management
  • Queues, domains, vMTA
  • Operational commands
MQ
MailerQMTA core
  • API integration
  • Queue monitoring
  • Server statuses
  • Operational control
PG
Prometheus · Grafanaobservability
  • Metrics scraping
  • Dashboards
  • Alerts
AI
AI Coreintelligence
  • Infrastructure planning
  • Audit and recommendations
  • Risk scoring
Who it's for

An infrastructure back office for six kinds of teams

From owners of their own email service to in-house email teams — wherever there are more MTA nodes than one head can hold.

SaaS Owners

Owners of Protosender-based services

Problem

Behind a commercial email service sits infrastructure that needs managing every day.

Solution

Central MTA control, warmup, monitoring and export of ready pools straight into Protosender.

Agencies

Email agencies

Problem

Many clients means many domains, IPs and SMTP sources scattered across projects.

Solution

A single registry of the whole infrastructure and per-client monitoring from one panel.

Deliverability

Deliverability specialists

Problem

It's hard to quickly see why delivery dropped: plenty of data, all in different places.

Solution

Queue metrics, errors, latency and AI audit — all diagnostics on one screen.

Integrators

PMTA / MailerQ integrators

Problem

Hand-edited configs across dozens of servers, no versioning, no change control.

Solution

Config templates, diff, validation and rollback — infrastructure managed as code.

Micro-ESP

Micro-ESP teams

Problem

Running your own email service demands constant infrastructure work, and the team is small.

Solution

MTA-Hub takes over the routine and becomes the service's infrastructure back office.

Internal Teams

In-house email teams

Problem

Transactional and marketing flows demand visibility, roles and access control.

Solution

Monitoring, alerts, role-based operations and AI analysis across the company's whole infrastructure.

Status: in active development

MTA-Hub is being built right now — and that's your chance

We are building MTA-Hub right now: the platform you saw above is getting ready to launch. Sales haven't started yet — which means early subscribers have the advantage.

🎁 Before sales open, subscribers of our Telegram channel will get a special price on MTA-Hub. The release announcement and module deep-dives land there too. Subscribe now so you don't miss the discount.

Subscribe to @master_mail Telegram channel: @master_mail
Sales launch announcement — first Special price for subscribers Module deep-dives as they ship