alexis.jeuffrain
All work

● Case study · CRM platform

OPTIMA

A modular CRM, and AI quotes a human signs off. Every small business has its own trade, and its tool has to speak it. One shared core, modules for each trade, and generative AI kept inside a space it cannot leave.

Client
Burst IT
Role
CPO, product and design lead
Team
2 developers, 2 product, 1 sales
Research
Seven industries
Status
In use, to be distributed
Years
2025–2026
Status
Shipped
OPTIMA screens laid out as a map: dashboard, contacts, tasks, calendar, quote, catalogue

Context

Every CRM was built for someone else's industry.

At Burst IT we sold fast, adaptable software to small and mid-sized companies. Nearly all of them ran on a CRM, and every CRM on the market was tuned for a different trade.

Research

Seven industries, one question: what is actually common?

I interviewed teams in laboratory testing, hospital maintenance, transport and delivery rounds, school supplies wholesale, e-commerce, prospecting and tax advisory. Two companies in the same sector rarely needed the same tool, because their stakes were different.

Architecture

One core, modules on top.

● The call

What belongs in the shared core, and what stays a module?

ConsideredOne configurable productSettings for every case. Grows heavier with each client.
ChosenShared core, trade modulesContacts, data and ease of use for everyone. Each trade gets its own module.

The core holds what everyone needs: contacts, customer data, a tool that is easy to pick up. Everything that expresses a trade is a module on top. OPTIMA also runs Burst IT itself: invoicing and payouts between clients, freelancers and partners.

The shared core

One dashboard, composed by the trade.

The dashboard every client sees first: leads, revenue, tasks, calendar. Customise opens a panel of widgets; a trade module adds its own, here the field rounds of the lab module, and the page reflows. That is how a module shows up without a new screen.

In numbers

7industries interviewed before the first screen
2trades rebuilt end to end in code: lab testing and school supplies
2themes, dark by default, light as its mirror
1human review before any quote leaves

Case 1 · A call becomes a quote

Dictated on the call. Fixed by a person. Sent.

The model drafts the lines from the catalogue only. A missed quantity is corrected, an unheard item removed, an unknown request resolved by a person. Nothing leaves before it is read.

Case 2 · The kit leaves, the stock follows

Assigned on the map. Reordered before it runs out.

OPTIMA ranks three technicians on drive time, skills, free time and kit; the planner assigns and the van leaves. In the depot, the slot that runs low is reordered, and the stock goes up when the pallet is scanned in.

Case 3 · On site, the office sees everything

The tablet in the field, the record at head office.

A count fails: the tablet flags it, the technician checks the seal and measures again, and the office sees both. Signed on the tablet, the report writes itself, and the room’s next requalification moves six months on.

Case 4 · Another trade, same core

A school supply list becomes an order, then a delivery round.

The list is read against the catalogue and multiplied by the pupils; two lines wait for a person. On the real roads of Lyon the rounds are optimised: about 30 % fewer kilometres, every school served outside class hours.