CAPTURE · CUSTOMER INTELLIGENCE
Customer Intelligence
One record per customer: what they did, what they agreed to, and when and where to reach them. Six condition families, the audience size before you save, Champions to Lost maintained for you.
Where teams start
All nine goals Three outcomes most teams pick first.
Start with the customer outcome that matters most to your team. Each goal page shows the capability behind it, the prerequisites and a journey you could ship in the pilot.
Journeys
Recover stalled journeys
The application left at the documents step, the checkout never paid, the sign-up that stopped. A rule catches what did not happen, one message follows on the channel they answer, and the journey stops when they finish.
Read the goalDelivery
Reduce avoidable communication cost
See reach and estimated cost before a send, keep each traffic type on its own route, and let fallback advance on a carrier failure rather than a timer. Your own provider contracts stay in place.
Read the goalAnalytics
Understand communication outcomes
Every message keeps its route, attempts and outcome on one record with the behaviour that followed. Report what a campaign was observed to change, what your attribution rules credit, and what a comparison supports.
Read the goalStalled checkoutJOURNEY
STSTARTReached checkout, never paid
WASENDWhatsApp with two buttons
WAITFor a reply or a button tap
REPLIED
SENDThe checkout link
SILENCE
SENDOne SMS, then stop
✓Every run recorded, node by node
Journeys your business team builds, that respond to the customer
24 building blocks. No release cycle. Every run explained.
A journey variation used to restart a sprint. Now it is an edit, made by the team accountable for the result.
- Send on any channel, or place a call, as a step in the same graph.
- Wait for a reply, a button tap or a keypress. Branch on what the customer chose, with a path for silence.
- Split an audience up to five ways and compare, inside the journey.
- Check and record consent in the graph, where a reviewer can see it.
- Call a system with a REST API mid-journey and branch on its answer, once your technical team has connected it and mapped the fields.
Every run is recorded node by node, with what the journey knew when it decided.
The platform
Platform overview One platform over one dataset.
ASK · ASKLYTICS
Analytics & Asklytics
Ask in plain language and get a live answer, checked before it runs. Boards, funnels, metric trees and sessions over one dataset, so the tool that finds the problem also tells you whether you fixed it.
ACT · STORIES
Stories
Full-screen, tappable content in your live app. No release.
ACT · JOURNEYS
Journey Automation
24 blocks. Waits for the customer, branches, calls your systems.
ACT · SIX CHANNELS
Messaging & Delivery
Six channels, one routing layer, one record. Push on our own network.
ACT · CAMPAIGNS
Campaigns
Reach and cost known before you press send. Each contact's local time.
MEASURE · CONVERSIONS
Measure · Conversions
Define which events count as a conversion and what each is worth. Every campaign and journey is measured against them, against a baseline, not opens. The result lands back as an event, and the loop runs again.
AI · BUILT IN, AND YOURS
AI Agents
Four agents, each with a named job, plus a Story generation pipeline. Every proposal shows its working; you accept or reject each item.
BUILT IN
AGENT 01Predictive SegmentationProposes audiences from your real data, each with a reason and a size, so the audience is a decision, not a guess.AGENT 02Copy WriterDrafts in your brand voice, ready for each channel, checked before it can send.AGENT 03Template ComplianceChecks every message against each channel's rules and your own blocklist before it goes live.AGENT 04Campaign InsightsReads the results and says what worked, what didn't, and the next three moves.
YOURSOPEN STANDARD · MCPBring your own agents. Connect a compatible agent through xNotify's open-standard MCP server and it can work the platform through the tools you scope to it. Agents that send messages or change external systems do so only with the permissions you grant, and every call is logged.
- Connect your own agents
- 112 tools your agents can use
- You set the scope
- Rehearse with a dry run first
- Every action logged; no delete tools exposed
You hold the controls
Change your channel order. The next message follows it.
When a carrier degrades at 2pm on campaign day, the fix is yours to make, not a vendor's ticket.
- Reorder channels for any message type and the change takes effect the moment you save. No deploy, no reload.
- Stop, resume or cancel a bulk send or a campaign while it is running.
- See the cost and the per-channel reach of a campaign before you press send.
- Keep one-time passwords on their own lane, with their own route and dispatch rate, so a marketing burst does not queue in front of them.
CHANNEL ORDER
One-time passwordsMarketingAlerts
1WhatsApp↑ MOVED UP
2SMSCarrier degraded
3Voice
4Email
✓Saved. The next message follows this order.
Effective on save
NO DEPLOY · NO RELOADBuilt for the buyers with the most to protect
Your data in its own database. Your deployment inside your own walls, if that's the rule.
- Each customer's operational data lives in its own databases; analytics events sit in a shared store partitioned by customer, with every query scoped to one tenant. A query on the wrong connection returns nothing, not someone else's records.
- The same platform runs as a shared service, a dedicated instance, or an on-premise, in-country installation.
- eOcean holds ISO 27001 and ISO 9001 at organisation level. A separate security white paper covers controls and independent penetration-test results.
ONE DATABASE PER CUSTOMER
QUERYon customer B's connection
CUSTOMER AReturns nothing
CUSTOMER BOwn database
CUSTOMER CReturns nothing
SHARED SERVICEDEDICATEDON-PREMISE, IN-COUNTRY
The activation gap
Your warehouse knows what happened. It doesn't know what's about to.
The investment landed on storage and reporting. The last mile, from insight to a customer, was never built.
Three queues stand between an observation and a customer contact. The person with the question raises a request and waits days for an answer. The answer becomes an engineering project: a spec, a sprint, a release. And once the journey is live, every correction runs through somebody else's queue again.
So teams ship few journeys and change them rarely. That is the opposite of what behavioural data rewards.
OBSERVATION TO CONTACTTODAY
Observation
Question to answerQUEUE 1
Waits days
Answer to journeyQUEUE 2
SPECSPRINTRELEASE
Every correctionQUEUE 3
Somebody else's queue, again.Customer contact
WITH ONE LOOP
ObservationCustomer contact
Same team, same afternoon.After the demo: an optional pilot
One use case. Six to eight weeks. A number you agreed before week one.
A demo comes first. If it is worth going further, the pilot is one journey, shipped and measured, not a migration and not a channel consolidation programme.
How the pilot worksCAPTUREASK ACTMEASURE
YOUR TEAM DRIVES →
Weeks 1–2
- What happens
- Turn on event capture. Agree the baseline metric.
- Exit condition
- Events arriving, schema visible, baseline recorded.
Weeks 3–4
- What happens
- Build audiences. Ask real questions.
- Exit condition
- Your team is querying without help.
Weeks 5–6
- What happens
- Build and test the journey. Dry-run and preview.
- Exit condition
- Journey approved, execution traces reviewed.
Weeks 7–8
- What happens
- Run live against the audience. Measure.
- Exit condition
- Result compared to baseline, decision made.
Your business team drives from week 3. Scope and timing are agreed around your integration and approvals.
If it fails to beat the baseline, that is a real answer, and worth having in eight weeks rather than after a platform commitment.