Observe · Hypothesize · Test · Learn

The scientific method,
at the speed of AI.

Your AI assistant can run a thousand experiments — but forgets everything between sessions. Apex captures what your team learns, structures it into testable beliefs, and compounds it into every future decision. Every cycle makes the next one smarter.

Free forever · No credit card required · Upgrade when you're ready

+43%

p < 0.05

Δ revenue

3 signals

apex — mcp

> "We close 2% of enterprise trials. Where are we

losing the other 98%?"

Scanning 14 months of deals across 312 accounts...

⚠ Found it. Your lead scoring is broken.

You’re prioritizing big logos, but mid-market accounts

close at 2.8x the rate — and you’re ranking them 40% lower.

✓ Fix: re-score leads by who actually buy, not who

looks impressive. Predicted impact: +$2.4M/quarter.

✓ Test: split 1,200 leads between old vs. new scoring

for 21 days. Measure real revenue, not vanity metrics.

✓ Running now — you’ll have results with full confidence.

This is your 8th correction. Each one made the next

one smarter. Your system is 6.2x better than day one.

+$2.4M

Revenue Impact

8th

Corrections

91%

Confidence

6.2x

Compounding

+43%

p < 0.05

Δ revenue

3 signals

apex — mcp

> "We close 2% of enterprise trials. Where are we

losing the other 98%?"

Scanning 14 months of deals across 312 accounts...

⚠ Found it. Your lead scoring is broken.

You’re prioritizing big logos, but mid-market accounts

close at 2.8x the rate — and you’re ranking them 40% lower.

✓ Fix: re-score leads by who actually buy, not who

looks impressive. Predicted impact: +$2.4M/quarter.

✓ Test: split 1,200 leads between old vs. new scoring

for 21 days. Measure real revenue, not vanity metrics.

✓ Running now — you’ll have results with full confidence.

This is your 8th correction. Each one made the next

one smarter. Your system is 6.2x better than day one.

+$2.4M

Revenue Impact

8th

Corrections

91%

Confidence

6.2x

Compounding

How It Works

The scientific method,
built into your growth stack.

Observe what happens. Form a hypothesis. Test it. Learn from the result. Repeat. Apex automates this cycle and compounds your team's intelligence over time.

Choose your depth — same system, three levels of detail

You describe what to test

Tell Apex what you want to try. It writes the experiment into your code, deploys it through your pipeline, and measures real business outcomes — not just clicks.

It learns from the result

Every experiment teaches the system. It remembers what worked, what didn’t, and updates everything it knows — automatically.

It tells you what to do next

Based on everything your team has ever learned, the system recommends the highest-impact action. Your 10th experiment is smarter than your 1st.

Each cycle makes the next one better. Knowledge compounds.

Three Interfaces

Three ways to build
intelligence.

Connect your AI agent via MCP for persistent intelligence. Drop in a script tag for zero-code experiments. Or pipe data from your existing stack through native sensors.

MCP Server

Your AI agent connects to Apex's belief graph, prediction calibration, and experiment engine. Every session starts from everything your team has ever learned.

# .cursor/mcp.json

{

"mcpServers": {

"apex": {

"command": "npx",

"args": ["@apex-inc/mcp-server"]

}

Install in Cursor Install in VS Code

Beliefs & predictions persist across sessions

AI suggests what to test next

Experiments ship as real code

Calibration tracks prediction accuracy

Script Tag

One line of HTML. Instant text experiments without touching your codebase. Perfect for marketers and non-technical teams.

Deterministic variant assignment

Automatic UTM capture & attribution

Form interception for conversions

Visual preview mode for QA

Text-based targeting — no selectors

Data Sensors

Native connectors to your analytics, CRM, ad platforms, and payment systems. OAuth in, data flows automatically.

OAuth connect — no engineering needed

Real-time data ingestion & sync

Unified event schema across sources

Cost data auto-mapped to campaigns

GA4

Ads

Sensors in. Decisions out.

Apex reads from the tools you already use and acts through built-in execution capabilities. Your stack stays. Apex makes it smarter.

Every growth team fights the same battles — disconnected signals, decisions based on intuition, and intelligence that evaporates between campaigns. Apex solves them regardless of your vertical.

Whether sales-led, product-led, or hybrid — close the loop from spend to revenue.

You can’t connect ad spend to closed revenue

reconciliation needed

Your analytics shows clicks. Your CRM shows deals. Nobody connects them. CAC and payback period are educated guesses.

Apex maps every event to channel spend and pipeline value. CAC and payback are always current.

Your experiments optimize for vanity metrics

Revenue

not clicks

A $50K enterprise demo and a free trial sign-up are not the same. Your tests can’t tell the difference.

Experiments scored by real business value per dollar. Pipeline, not pageviews.

Your team learns something every week — and loses it by next quarter

10×

faster time-to-insight

Insights live in dashboards nobody re-reads, experiments nobody references, and tribal knowledge that leaves when people do.

Apex turns every signal into institutional memory. The loop compounds — so the 50th decision is dramatically better than the first.

Key sensors

Google AdsLinkedIn AdsHubSpotGA4

Cost per weighted pipeline dollar

Same platform. Different verticals. The scientific method works everywhere. Your sensors, metrics, and experiments adapt to your business — the intelligence loop stays the same.

Pricing

Start free. Upgrade when the loop pays for itself.

The intelligence layer is free forever. When you're ready to close the loop — experiments, budget allocation, AI-scored decisions — you pay for the value Apex generates, not the tools it touches.

Free

Forever

$0/mo

The full intelligence layer. Install the SDK, connect your data, see weighted scores and timeline intelligence. No credit card. No time limit. Every user makes the system smarter for everyone.

Core scoring engine

Timeline intelligence

Identity resolution

Executive dashboard

Open-source SDK + all connectors

Belief graph (read-only)

2 active experiments

MCP server access

5 lifecycle communications with standard branding

Up to 2 seats

50K events/mo

When you're ready to close the loop

MonthlyAnnual

Save 20% with annual billing

Starter

Close the loop. Unlimited experiments evaluated by weighted business value.

$320/mo billed annually

Up to 3 seats · 200K events/mo · Unlimited projects

Everything in Free, plus

Unlimited experiments + manual evaluation
15 lifecycle communications tailored to your vertical
2 communication experiments per month with basic A/B testing
3 concurrent A/B tests with manual winner promotion
Mobile install attribution
Cross-platform identity (web + mobile)
Up to 3 seats · 200K events/mo

Pro

Optimize the loop. Budget intelligence, calibration tracking, and learning archive.

$1,200/mo billed annually

Up to 5 seats · 500K events/mo · Unlimited projects

Everything in Starter, plus

Full timeline intelligence (read + write)
Prediction calibration tracking
Blind mode for sales teams
Experiment history + learning archive
Scoring model drift detection
Offline conversion tracking via CAPI
CRM score writeback
CPWU tracking across channels
Email support
50 communications with your brand — no 'Powered by' footer
10 experiments with adaptive Thompson Sampling and up to 4 variants
10 concurrent tests with auto-completion at statistical significance
SKAN 4.0 conversion management
Cross-platform journey view

Every user makes the system smarter for every other user.

Apex aggregates anonymized patterns across all organizations — scoring configurations, experiment outcomes, model drift signals. The result: better default templates, industry benchmarks, and AI recommendations that improve with scale. This is the same network effect that made Stripe's fraud detection and Shopify's conversion benchmarks defensible.

nodes 0

connections 0

Templates

Default scoring models improve as more companies across verticals contribute data

Benchmarks

Compare your weighted performance against similar companies in your vertical

Recommendations

AI suggestions get better as the system learns what works across industries

Stop guessing.
Start compounding.

Every experiment your team runs builds institutional knowledge that compounds over time — and survives team turnover. Every day you wait is a day your competitor's system learns something yours doesn't.

Free forever · No credit card required · Set up in under 5 minutes

The scientific method,at the speed of AI.

The scientific method, built into your growth stack.

You describe what to test

It learns from the result

It tells you what to do next

Three ways to build intelligence.

Sensors in. Decisions out.

Sensors

Google Analytics 4

Google Ads

LinkedIn Ads

Meta Ads

HubSpot

Stripe

QuickBooks

Microsoft Clarity

Apex SDK

Salesforce

Mixpanel

Amplitude

PostHog

Segment

Actuators

Experiment Engine

Task Generation

Budget Allocation

Approval Workflows

Webhooks

Slack Notifications

You know these problems.