Cost
Up to 17×
more cost-effective
- Claude Sonnet 4.5Build it yourself−94%
- Claude Haiku 4.5Build it yourself−83%
- OpenAI GPT-5 miniBuild it yourself−49%
- Gemini Flash-LiteBuild it yourself · limited—
- Maestro 2.0Ready to use
Introducing Maestro 2.0
With Maestro 2.0, AI CRM Workforce for Retailers
Cleanse
Merged, fixed and clean, before any decision.
Enrich
Gives every SKU the context your data never had.
Remember
Household, habits and reasons, remembered.
KPI & Guardrails
You set the KPIs and limits. Hand over the ownership.
Skills
Retail expertise built in, deepened as you grow.
Decide & Learn
A 1:1 lifecycle for each customer, at any scale.
01Cleanse
Before any decision, duplicates are merged, broken fields cleansed and potential test accounts quarantined, so each one rests on one trusted record per customer
02Enrich
A SKU says little. It adds the context your data never had: who a product suits, when it is used, what replaces it and what it pairs with. So the right substitute, cross-sell and refill reach each customer, with no manual tagging
03Remember
It remembers who each customer is, not just what they clicked: their household, why they buy, who they buy for and how it all changes. Each decision starts from the whole person, and every outcome adds to it
04KPI & Guardrails
Tell it what the business needs and where the limits are. Every decision follows both
05Skills
Start with 100+ pre-built retail skills, deepen the ones you need, or build your own in plain language
06Decide & Learn
It makes a 1:1 decision for every customer and sends that intelligence to every surface: your website, CRM, app and more, and learns from the outcome
Follow what Maestro does for Charlotte, from a thin first record to the right message on every surface, and then for everyone else
One order, like about 40% of your database. Maestro enriches that thin record with context, so it knows who Charlotte really is, not just what she bought
Illustrative. Invented example data.
Maestro listens to every signal from its memory. Today it proactively picked Charlotte, and 21,304 other customers, to engage, and for each one it does the following, uniquely:
Illustrative. Invented example data.
Wherever she meets your brand, what she sees is written just for her, one to one:
Illustrative. Invented example data.
Autonomously.Every day.One to one.On every surface.Learning from every outcome.
Illustrative. Invented example data.
Case studies
Real retailers that handed Maestro the decision, and what it returned
How to start
Maestro starts from three things you already have. It builds the rest, and decisions leave through the platform you already send from
01What we need from you
02Maestro builds the rest
03Switch on
Maestro 2.0 in numbers
Cost
Up to 17×
more cost-effective
Model improvements
9.1×
revenue per decision

Maestro’s case study with Google
Skill Studio
Maestro comes with retail expertise built in. Add your own strategies in plain language, and Maestro thinks them through for each customer
100+ retail skills, ready to switch on
Video coming soonPre-built skills walkthrough
In the press
The latest announcements, partnerships, and milestones from Replenit
ofarmakopoiosmou.gr hires Maestro, Replenit's AI CRM Manager, to run 1:1 retention, recommendations and My Account pages, with no more software to manage and specific skills for agents to operate.
Read the Full AnnouncementMaestro, Replenit's AI Decision Engine, is appointed AI CRM Manager for Estée Lauder Companies EMEA, managing retention, cross-sell, replenishment, and engagement across the brand portfolio.
Read the Full AnnouncementBeko Corporate selects Replenit as its AI Decision Engine to power intelligent, individualized post-purchase customer engagement across connected appliances and lifecycle signals.
Read the Full AnnouncementMaestro starts on the stack you already run and makes it unnecessary over time, managing every customer’s lifecycle 1:1 from one memory. Your team leads. The AI operates
FAQ
How Maestro works, how it decides, and what it takes to go live. If you run multiple brands or business units, start with the Multi-Brand & Group tab
Replenit is an AI decisioning layer for retail. It sits between your data stack and your engagement systems, turns customer, product and brand records into a living memory, and decides what should happen next for each customer, continuously.
The output is execution-ready, not a report or a segment list.
Don't see your question, or want this scoped against your current stack and brand portfolio?
Talk to us