Skip to content

/ 01 — Category

The email delivery platform built for senders who actually need their email to arrive.

Email Delivery Platform is the infrastructure layer between your application and the mailbox providers — Gmail, Yahoo, Outlook, the rest. We run our own dedicated mail transfer agents — PowerMTA and KumoMTA, the engines professional senders use — as a European entity under EU jurisdiction, not a shared pool or a reseller of someone else's cloud. Two tiers run in parallel: a self-serve SaaS with an API for teams that need to start sending today, and a managed tier where we provision dedicated IPs, warm them, and operate deliverability on your behalf. Pick the track that matches your sending — not the one a vendor wants to sell you.

Managed delivery operating · 99.95% historical uptime

smtp · port 587 · TLS
220 mx.emaildeliveryplatform.com ESMTP ready
EHLO app.acme.io
250-STARTTLS
250-AUTH LOGIN PLAIN
250 SIZE 52428800
MAIL FROM:<invoice@acme.io>
250 2.1.0 sender ok
RCPT TO:<buyer@gmail.com>
250 2.1.5 recipient ok
DATA
354 start mail input
. . .
250 2.0.0 queued as 14a2f9c
QUIT
221 bye

A successful transactional handoff against our MX. Authentication checks, queue assignment, IP selection, and reputation routing happen behind it.

/ At a glance

  • 2 tiers

    SaaS self-serve and managed infrastructure run in parallel. Same backbone, different operational model.

  • ≥97%

    Inbox placement target on managed dedicated pools after warming completes, measured against a third-party seed list.

  • KumoMTA

    Primary MTA engine on managed. PowerMTA supported for migrations. Postfix where it fits.

  • EU entity

    Operated under EU jurisdiction. A US provider's EU region is residency, not jurisdiction — the CLOUD Act still reaches it. Five languages; DSGVO, RGPD, LGPD posture documented.

  • Per-stream

    Transactional and marketing on separate pools by default. A campaign incident cannot sink a password reset.

/ The distinction

Email delivery vs email deliverability — what is the difference?

Email delivery is whether the receiving server accepts your message at all — a binary outcome measured as the delivery rate: accepted, or bounced. Email deliverability is whether the accepted message reaches the inbox rather than the spam or promotions folder — measured as inbox placement, and governed by sender reputation, authentication, and engagement. A message can be delivered and still have poor deliverability: the server took it, then filed it in spam. The first is a property of the connection; the second is a property of your reputation. A platform has to operate both — and most of the work that matters happens in the second.

From your app to a person who reads it — two stages, not one 1 · Delivery Server accepts? 250 OK / bounce metric: delivery rate 2 · Deliverability Inbox or spam? reputation · auth metric: inbox placement Inbox Spam / Promotions read by a person delivered but in spam = good delivery, poor deliverability — the gap most senders never measure

The distinction matters because the two are bought and sold as if they were one thing, and they are not. A provider can report a delivery rate above ninety-nine percent — almost everything was accepted by the receiving servers — while a large share of those accepted messages quietly land in spam. The delivery number looks excellent and the deliverability is poor, and a sender watching only the first figure never sees the problem until replies and conversions fall off.

That is the layer Email Delivery Platform operates. Delivery is mostly solved by competent infrastructure: authenticated, well-connected mail servers that receivers accept. Deliverability is the ongoing discipline — dedicated IPs warmed to your sending curve, reputation watched daily across each receiver, authentication maintained, streams isolated so a marketing batch cannot sink a password reset. Running our own MTAs is what lets us operate the second layer rather than just report on the first, which is the difference between a platform that delivers your mail and one that gets it read.

/ 02 — The two tracks

What is the difference between SaaS and managed email infrastructure?

SaaS email platforms give you an API and a dashboard, then hand the operational responsibility back to your team. Managed infrastructure runs the operations on your behalf — same MTA engines underneath, very different working relationship. Most platforms force you to pick one philosophy. We run both because most sending portfolios contain both.

Track A

SaaS self-serve

Sign up, generate an API key, send. Monthly billing tied to volume. The SaaS tier exists for teams who need to ship email today and have a clear technical owner who can monitor sender reputation on their own.

You get

  • REST API and SMTP relay with documented SDKs
  • Shared sending pools plus the option to attach a dedicated IP from your dashboard
  • Webhook events for bounces, complaints, opens, clicks
  • Postmaster Tools and SNDS data piped into the dashboard
  • Pricing public, transparent, no overage roulette

Best fit

Teams sending under a few million per month. Internal engineering capable of reading placement data and acting on it. Use cases where the developer is the buyer.

See the SaaS tier

Track B

Managed infrastructure

Dedicated IPs provisioned and warmed by our deliverability team. Transactional and marketing on isolated pools. The operational work — monitoring, troubleshooting, reputation repair, incident response — happens on our side of the line.

You get

  • Dedicated IPs provisioned, warmed, monitored by a named engineer
  • Stream isolation by default — transactional, marketing, system messages on separate pools
  • KumoMTA or PowerMTA underneath, chosen as part of the architecture review
  • Direct contact with the team operating your pools — not a tiered ticket queue
  • Pricing decoupled from message volume — flat platform plus per-IP, no per-email tax

Best fit

Senders past the self-serve tier — millions per month, multi-stream, multi-tenant, or recovering reputation. Buyers who treat email as infrastructure, not as a marketing tool.

Request an infrastructure review

/ 03 — Technical foundation

The MTA layer: KumoMTA, PowerMTA, and where Postfix still fits.

KumoMTA is the primary engine on the managed tier — modern, Lua-scriptable, built by operators who ran high-volume email programs at scale and wanted an MTA designed for the work they actually do.

PowerMTA remains supported because some senders need to migrate without rewriting their entire send pipeline at once. Postfix shows up where the volume and the policy complexity make a commercial MTA unnecessary overhead. The choice is part of the architecture review — what fits the sending pattern, not what costs us less to operate.

On the SaaS tier the engine is abstracted away — you get an API. Underneath, the same KumoMTA cluster handles your mail. The difference is operational, not architectural.

lua · kumomta init.lua excerpt

production

-- Per-tenant egress shaping: each customer routes through its own pool
-- Transactional vs marketing isolation enforced at queue-construct time

kumo.on("init", function()
  kumo.start_esmtp_listener {
    listen = "0.0.0.0:587",
    tls_certificate = "/etc/ssl/edp/fullchain.pem",
    tls_private_key = "/etc/ssl/edp/privkey.pem",
    relay_hosts = { "127.0.0.1" },
  }

  -- Define per-stream egress sources
  kumo.define_egress_pool {
    name = "transactional-acme",
    entries = { { name = "acme-tx-ip-01" } },
  }
  kumo.define_egress_pool {
    name = "marketing-acme",
    entries = { { name = "acme-mkt-ip-02" } },
  }
end)

-- Route messages to the correct pool based on X-EDP-Stream header
kumo.on("smtp_server_message_received", function(msg)
  local stream = msg:get_first_named_header_value("X-EDP-Stream")
  local tenant = msg:get_first_named_header_value("X-EDP-Tenant")
  msg:set_meta("egress_pool", stream .. "-" .. tenant)
end)

A real excerpt: per-tenant pool definition with stream-aware routing. Same pattern runs across every managed customer. The Lua script is the source of truth — not a UI abstraction over a black box.

/ 04 — Architecture

What happens between your application and the recipient's inbox?

The diagram below traces the path of a message from the moment your application hands it to us until the receiving mailbox provider returns a final response. Every step has receivers, reasons it can fail, and operational signals worth monitoring. Most platforms hide this layer behind a green checkmark on a dashboard. We treat it as the work.

Your application Generates message, sets stream header, calls our API or SMTP Edge ingestion Auth check, rate limiting, tenant identification KumoMTA queue Pool selection by stream, DKIM signing, receiver-rate shaping Dedicated IPs Warmed, monitored, stream-isolated, reputation tracked Mailbox provider Gmail, Yahoo, Microsoft 365, corporate gateways — Operated by Email Delivery Platform — Authentication, queue management, pool selection, warming, monitoring, incident response Postmaster Tools · SNDS · Yahoo Insights feedback / 01 / 02 / 03 / 04 / 05 Message lifecycle: application to mailbox Stage 01 is yours. Stages 02 through 04 are ours on managed. Stage 05 is the receiver and is never ours.

Where the work that matters lives

Stages two through four are where deliverability is made or lost. Edge ingestion authenticates and rate-limits to prevent abuse from compromised customer accounts. The MTA queue chooses which IP pool a given message should leave from, signs it with the right DKIM key for the From domain, and shapes the egress rate to match what each receiver currently tolerates. Dedicated IPs carry the long-term reputation that determines whether the same authenticated message lands in the inbox or in the spam folder.

The receiver side, stage five, is never anyone's to control. We can influence what they see — clean authentication, low complaint rate, engaged recipients — but we never make the placement decision for them. Anyone who claims otherwise is selling you optimism, not infrastructure.

The feedback loop is where operations live

The dashed arc at the bottom of the diagram represents the daily feedback work. Postmaster Tools gives reputation data from Gmail with a twenty-four to forty-eight hour lag. Microsoft SNDS reports color-coded sender ratings and complaint rates from Outlook properties. Yahoo Insights, which launched in October 2025, finally fills the third major gap. Reading these dashboards every morning is the work that catches reputation drift before it becomes a placement collapse.

Two to seven days is typically the window between the first reputation signal showing yellow and the receiver flipping placement to the spam folder. Caught early, the recovery is conversational with the receiver and routine on our end. Caught late, it becomes a multi-week remediation with revenue impact. The work is monotonous and the value compounds.

/ 05 — Fit

Who should use Email Delivery Platform?

We have written this section in the form of who fits, who does not, and why. The honest version of the answer matters more than the marketing one — if you do not fit, our tools will not save you, and the diagnostic conversation should happen before money changes hands.

Fits

SaaS platforms past the free tier

Multi-tenant products sending transactional and lifecycle email across many customer bases. One bad tenant on a shared pool damages every neighboring tenant — stream and tenant isolation become product-quality concerns, not marketing concerns.

Fits

High-volume publishers and ecommerce

Newsletters past one million sends per month. Ecommerce operators with peak campaigns that look like spam to mailbox providers unless warming and pacing are handled deliberately. Both benefit from dedicated pools and per-campaign monitoring.

Fits

Agencies running clients' programs

Per-client pool isolation, white-label reporting, the ability to quarantine one client's deliverability incident before it touches the others. Agencies usually arrive here after one shared-pool blowup convinced them the math does not work.

Does not fit

Hobby projects and small SaaS

Under ten thousand sends per month, the SaaS tier of Postmark or Resend will be cheaper and operationally adequate. We will say so on the discovery call rather than upsell you into a managed plan that costs more than it returns.

Does not fit

AWS-native, SES-capable teams

Amazon SES at $0.10 per thousand is structurally cheaper than any managed provider if you can operate it yourself. If your team has the bandwidth to stitch together CloudWatch, SNS, SQS, and the suspension-tolerant retry pattern, SES is the right call.

Does not fit

List-quality problems

If your placement problem traces to old, unengaged, or improperly acquired addresses, no platform migration will fix it. ZeroBounce, NeverBounce, or similar validation services will help you faster than we will. We may still take you on once the list is cleaned, but the order matters.

/ 06 — Compare

How does this compare to SendGrid, Mailgun, Amazon SES, and Postmark?

The short answer is that all four are SaaS — the long answer requires more nuance, and we have written it page by page. Each comparison has a section that says when the competitor is the honest right answer, because the audience reading these pages can tell the difference between a fair comparison and a sales document.

Vendor Model Their edge Where we sit Deep dive
SendGrid SaaS, Twilio-owned Twilio integrations, mature SDKs, enterprise compliance certificates Operational layer included on managed; transparent per-IP pricing Read →
Mailgun SaaS, Sinch-owned Inbound parsing API, validation API, developer documentation Per-email pricing replaced by flat platform + per-IP; Outlook posture stronger on managed Read →
Amazon SES Infrastructure, AWS-native Cheapest per-email on the market at $0.10/1K; native AWS integrations The dashboard, the warming, and the ops you would otherwise build around SES, included Read →
Postmark SaaS transactional-only Top-of-field shared inbox placement at ~83% in independent tests Transactional and marketing on one platform without the two-platform tax Read →

Each comparison page ends with a When to stay with [vendor] section. The answers are not the same across pages, and they are not always us.

/ 07 — Operations

What does deliverability operations actually mean?

On the managed tier, this is the work that happens on our side of the line rather than yours. It is what distinguishes managed infrastructure from a SaaS API. The list is not exhaustive — it is the visible portion of the work that exists whether anyone is doing it or not.

  1. 01

    IP provisioning, warming, and graduation

    Acquire dedicated IPs from clean ranges. Run the four-to-six week graduated warming process with engagement-tier cohort sequencing. Promote each IP into full production only when receiver-side signals confirm the reputation is stable.

  2. 02

    Authentication setup and ongoing alignment

    SPF, DKIM, and DMARC configured with strict alignment between the From domain and the signing identity. DMARC policy graduated from p=none to p=reject on the timeline your audit cycles allow.

  3. 03

    Daily monitoring across Postmaster Tools, SNDS, Yahoo Insights

    Reputation signals from Gmail, Microsoft, and Yahoo dashboards read every morning. Trend deviations trigger investigation before they become incidents — the lead indicators precede the placement drop by two to seven days when caught early.

  4. 04

    Stream isolation and queue management at peak

    Transactional, marketing, and system messages route through separate pools by default. A campaign that spikes complaint rate cannot drag down a password reset. Queue depth is shaped by receiver-specific rate limits we keep current.

  5. 05

    Bounce processing and complaint handling

    Bounce categorization, automatic suppression of hard bounces, complaint feedback loops with the major receivers wired in. Soft-bounce retry curves tuned per receiver rather than one global policy.

  6. 06

    Incident response with a named engineer

    When placement drops or a receiver returns 421 throttling, you get the engineer running your pools — not a tier-one ticket queue. Communication happens in your channel of choice, on the timezone you operate in.

/ 09 — Commitments

What we will not promise, and why that matters.

Email infrastructure is one of those categories where the language of sales pages has drifted far from the language of the receivers actually making the placement decisions. We have written this section to mark the line between what we can commit to and what nobody honest can commit to.

We will not guarantee a percentage inbox placement number.

Placement is decided by the receiving mailbox provider on signals that include things we cannot see — historical recipient behavior, content classifier updates, cohort-level reputation shifts inside the receiver. Anyone selling a placement percentage as a guarantee is either misunderstanding the system or counting on you to. We commit to specific operational practices that make excellent placement achievable, and we report seed-list data weekly without selective framing.

We will not take an account whose list quality is the actual problem.

If the discovery call reveals that the placement problem traces to old, unengaged, or improperly acquired addresses, no migration to our platform will fix it. The first conversation may end with us recommending a list-validation pass before we resume the discussion. Taking on a damaged-list account and watching the reputation collapse on our IPs would be bad for both of us — we would rather lose the deal than pretend.

We will not invoice for promised volume you did not send.

The managed pricing model is flat platform fee plus per-IP cost. It is not per-email, and it is not a volume-tiered SaaS plan with commit thresholds that turn into overage on the next bill. If your sending shape changes mid-quarter — campaign cadence drops, a vertical pauses, a product launch slips — the cost on our side does not punish you for the variability. The cost behaves the way operating costs should: stable when your operations are.

/ 10 — Onboarding

What the first ninety days look like on managed.

A managed engagement is not signed today and producing inbox results tomorrow. The first quarter is mostly about provisioning, warming, and getting the deliverability baseline measured. Knowing that timeline before committing avoids the disappointment that ends most platform migrations badly.

  1. W1–W2

    Discovery, scoping, contracts

    Architecture review of your sending shape, dedicated IP count proposed, stream allocation discussed, contract executed. Authentication audit of your current SPF, DKIM, and DMARC happens in parallel — most engagements surface at least one misalignment here.

  2. W3–W4

    Provisioning and dual-send

    IPs provisioned from clean ranges. DNS records added. Dual-send mode begins — a small percentage of mail leaves through our pools alongside your existing provider so reputation is being built on our side before traffic ramps.

  3. W5–W8

    Warming and graduated migration

    Engagement-tier cohort sequencing. Volume doubles every two to three days as long as receiver-side signals stay clean. By week eight, most engagements have full transactional volume routed and marketing volume at fifty to seventy-five percent.

  4. W9–W12

    Stabilization and the first quarterly review

    Full volume on dedicated pools. First quarterly business review with placement data, complaint trends, receiver-by-receiver performance. Operational rhythm established for the rest of the engagement.

/ 11 — FAQ

Questions worth answering directly.

The questions buyers actually ask before signing — answered plainly, with the parts we cannot promise marked as such.

What is an email delivery platform?

The infrastructure layer that accepts mail from your application, signs and routes it through dedicated or shared IPs, and works to land it in the inbox of the receiving mailbox provider. The category sits between your application code and the mailbox providers themselves. Done well, it absorbs the operational work that would otherwise sit on your engineering team — IP warming, authentication, reputation monitoring, complaint handling, bounce processing, and queue management at peak.

What is the difference between SaaS and managed email infrastructure?

SaaS email platforms give you an API and a dashboard; you keep the operational responsibility for reputation, deliverability monitoring, and incident response. Managed email infrastructure inverts that: dedicated IPs are provisioned and warmed for you, deliverability operations are run by a named team, and the work that SaaS leaves to you becomes the service itself. Email Delivery Platform offers both tracks under the same brand because most real sending portfolios contain both kinds of need.

Who is Email Delivery Platform for?

The SaaS tier suits teams sending under a few million emails per month with a clear technical owner who can monitor placement themselves. The managed tier suits teams sending high volume across multiple streams, teams whose deliverability has degraded on a shared pool, and teams whose sending matters enough that one missed campaign costs more than a year of managed engagement.

What MTA does Email Delivery Platform use?

KumoMTA is the primary engine on the managed tier — modern, Lua-scriptable, designed for high-volume operators by people who ran high-volume email programs. PowerMTA is supported for migrations from legacy stacks. Postfix shows up where the volume and complexity make a commercial MTA unnecessary overhead. The choice is part of the architecture review, not a feature flag.

Do you guarantee inbox placement?

No, and any vendor who does is misleading you. Inbox placement is determined by the receiving mailbox providers based on signals you partially control (authentication, complaint rate, engagement) and signals you do not control (recipient behavior, mailbox provider algorithm changes, content classifier updates). We commit to specific operational practices that make excellent placement achievable, and we report the data honestly. We do not commit to a number we cannot enforce.

When should I not use Email Delivery Platform?

If you send under ten thousand emails per month, the SaaS tier of Postmark, Resend, or a similar transactional provider will be cheaper and adequate. If your sending is entirely on AWS and your team can operate Amazon SES with confidence, SES is structurally cheaper at scale and we will say so. If your problem is list quality rather than infrastructure, an email validation service will help you faster than a platform migration. We have written that advice into our comparison pages explicitly.

/ Next step

Two doors. Pick the one that fits.

You can start sending on the SaaS tier in fifteen minutes — signup, API key, send. Or you can request an infrastructure review and we will scope what a managed engagement would look like for your sending. Both paths are honest about where the other one fits better.