English | Customer Support | Contact Sales
Log in | Start free with Google | About
AI Agents · Inside the product, on your data

AI Agents That Act, Not Just Report

Lux AI, the in-app assistant, answers questions about your traffic, forms and leads — with live company data behind every reply. An onboarding agent configures the platform for your business. Research agents analyse websites and prospects. All inside Digenio AI, all with per-company memory.

Talk to the agent in the demo Start for free
Lux AI data assistant Onboarding agent Research agents Per-company memory & isolation
area.digenio.ai/v2
agent online
Lux AI logo
Lux AI ● assistant · knows your company data
You
How did the spring-launch campaign do this week?
Spring-launch brought 312 sessions (↗ 18% vs last week) and 21 form submissions — your best source after organic. Landing page /spring-offer converted at 6.7%, above your 4% target.
sessions ↗ 18% leads: 21 conv 6.7% source: utm/spring-launch
You
Which pages should I fix first?
Lux AI is reading your data…
Live context per reply: traffic sources · form conversions · UTM campaigns · real-time users · leads · smartpages
The old way

Dashboards Tell You What. Nobody Tells You What to Do.

You don't need more charts to interpret. You need an assistant that already knows your business and answers, in plain language, with your numbers.

Insight without action

The chart says conversions dropped. Then what? You open a spreadsheet, rebuild the context, and start guessing. The gap between seeing a number and acting on it is where the week goes.

Chatbots that know nothing

Generic AI assistants can write an email, but ask "how did my campaign do?" and they hallucinate or shrug. Without your live data behind the answer, an assistant is a parrot with a UI.

Setup that never ends

New tools arrive empty: no goals, no KPIs, no context about your business. Weeks later they still don't know what "good" looks like for you — so their reports average everyone and help no one.

What you can do

Agents That Onboard, Research and Answer

Available today inside the product — clearly separated from what's on the roadmap.

Lux AI: ask your data anything

In-app assistant that answers questions about traffic, campaigns, forms, leads and smartpages — grounded in your live company data, never generic. Per-company session memory keeps the conversation continuous.

● Live today

Onboarding agent

A guided chat that configures the platform for your business: goals, conversion events, market, revenue model, sales cycle and KPIs. Done in minutes, not a weekend of settings pages.

● Live today

Website analysis agent

Point it at any website and it builds a structured business profile — market, positioning, tone, goals — that feeds Company Intelligence and every recommendation across the platform.

● Live today

Prospect research agent

For sales teams: generates prospect intelligence reports — who they are, what they sell, where the angle is — before the first call. New business qualification without the manual digging.

● Live today

Your brand, your agent

Each agent carries your identity and voice — tone, language and role are configured per company. Multi-provider model routing picks the right AI model for each job, and every action is audit-logged.

● Live today

Coming next

Dashboard Data RAG — ask deep-dive questions across full dashboard history; Perplexity-powered external research; autonomous product demos; agentic integrations that act on your stack.

Roadmap
Under the hood

Every Answer, With Full Context

No stale prompts, no manual setup: the system prompt is rebuilt fresh for every message.

1

You ask

A message in the Lux AI chat — "how did the campaign do?", "which form converts best?", "what changed since yesterday?"

2

Context assembles

The agent service merges your identity, mission, business profile, voice — and live project data: traffic sources, conversions, engagement, real-time users, UTM campaigns, leads, recent activity.

3

The model answers

A multi-provider router sends the enriched prompt to the right AI model for the job — Anthropic, OpenAI, xAI and more, switchable per agent.

4

Memory keeps it going

Session history is stored per company — follow-up questions just work. Every interaction is audit-logged, with usage rolled up per agent.

In the product

Three Agents, One Assistant Bar

A representation of the v2 agent experience — from first setup to daily questions.

Onboarding chat
Business goal · lead gen
set
Conversion event · form submit
set
Market & language · UK · EN
set
Revenue model · B2B retainer
set
KPI targets · in progress…

The agent asks; the platform configures itself. No settings maze.

Ask, grounded in data
You
Why did traffic drop yesterday?
Total sessions fell −12%, but only on direct traffic — the spring-launch email ended Tuesday. Organic (+9%) and referral (+4%) both grew. No tracking issue detected.
Want me to compare channels for the full month?

Real sources, real numbers — the reply cites your data, not a guess.

Agent activity · audit log
14:02:11 [flowy] context built · 9 data blocks · 384 tok
14:02:12 [router] model: primary · provider ok
14:02:14 [flowy] reply delivered · grounded ✓
13:47:03 [onboarding] step 4/7 answered · revenue model
13:41:52 [research] website analysis queued · acme.co.uk
13:41:55 [guard] guideline applied · no external send
13:41:58 [research] profile stored · company_ai_profile

Every agent action logged — usage, context and guardrails visible.

Data & integrations

Your Data In, Grounded Answers Out

Context sources
  • Business profile & onboarding answers
  • Live analytics: sources, conversions, engagement
  • Real-time users & UTM campaigns
  • Forms, leads and smartpages summaries
Agent runtime
  • Per-company session memory (chat history)
  • Identity, voice & tone per company
  • Multi-provider model routing
  • Audit logging & usage rollup
Destinations
  • Answers in the Lux AI chat, inside the product
  • AI profile feeds Company Intelligence
  • Insights surfaced in your dashboards
  • Lead routing via Sphere CRM connector
FAQ

Questions, Answered

Today: the Lux AI data assistant (live answers grounded in your analytics, forms, leads and smartpages), the onboarding agent, the website analysis agent and the prospect research agent — all with per-company memory and audit logging. Roadmap, labelled clearly: Dashboard Data RAG (deep questions across full history), Perplexity-powered external research, autonomous demos and agentic integrations. We don't sell futures as presents.
No. Your data is never used to train models. Each company's data lives isolated on its own shard via the Backbone registry, agent memory is scoped per company, and every agent interaction is audit-logged. When the agent answers, it reads your data through the same permissioned services the product uses — nothing more.
Yes — identity, role, voice and tone are configured per company, and the agent answers in your working language. Because the system prompt is rebuilt from your profile and onboarding answers on every message, a change to your business context is reflected in the very next reply.
Not one — several. A model router sends each request to the right provider (Anthropic, OpenAI, xAI and more), switchable per agent and per use case. When a better model lands, it's a routing change — not a re-platform. You get the best answer available, not the only model we own.
Before every reply, the agent service assembles a fresh context: your business profile and onboarding answers, plus live data from Dashboards — traffic sources, conversions, engagement, real-time users, UTM campaigns — and your forms, leads and smartpages. No copy-pasting numbers into a chat window. The agent already did the reading.

Stop guessing.
Start knowing.

See exactly which marketing drives revenue —
from first click to closed deal.

Start free trial Start free with Google See how it works
Free 7-day trial
GDPR ready