Skip to content
team banzai

tech demos

Tech demos

First we build it. Then we tell it.

At Team Banzai we build two kinds of things we can show: our own products already in use, like TeachThemAll and HoyEstudia, and these experiments: technical demonstrations built on real, public data, published with the wins, the misses and the caveats. We keep publishing these internal tests on our site, so you can see the kind of work we can do for our clients.

Every experiment is different: a different sector, problem and technology. All of them can be verified, and the rules of each one are set before looking at the outcome: the sealed records can be downloaded and checked.

Several of these things, until recently, were not worth automating: building them cost more than they saved. That has changed, and that is why we try them.

Sector
Problem
Technology
Capability

Tomorrow's electricity price, predicted and sealed every day

Energy Forecast Statistics Bilingual

Enter →

Anticipating car recalls by reading driver complaints

Automotive Early warning Text analysis Bilingual

Enter →

Predicting the discount and the no-bids in public procurement

Public sector Forecast Statistics + classical ML Bilingual

Enter →

Predicting which files a fix will touch before it's written

Software Forecast Repository mining Bilingual

Enter →

The same map, two questions, opposite answers

Software Audit Dependency graph Bilingual

Enter →

Advising on a never-seen ticket with years of past incidents

Support Advise Generative AI Bilingual

Enter →

Flagging which changes deserve a second look before merging

Software Early warning Risk model Bilingual

Enter →

Reading a company's code before you buy it

Investment / M&A Audit Python + git, classical statistics, no LLM Bilingual

Enter →

The supermarket as a graph

Food retail Audit Graphs Bilingual

Enter →

Knowing which claims will cost money the day they arrive

Support Prioritise Classification Bilingual

Enter →

Matching payments to contracts with the key hidden, without the reference that links them

Public sector Reconcile Deterministic cascade + language model Bilingual

Enter →

Relating every medicine in Spain and showing what its official label says

Health Relate Knowledge graph Bilingual

Enter →

Rebuilding a software's documentation by reading its code, exam included

Software Document Generative AI + deterministic checker Bilingual

Enter →

Solving a real school's timetable with a free optimizer

Education Plan Constraint optimisation Bilingual

Enter →

Shaving one point off a hospital's best known roster

Health Plan Constraint optimisation Spanish

Enter →

Under construction

Planning: timetables, rosters and workshops

The results of Telar, our planning engine, on public problem collections. They will be published once the maintainers of each collection verify and list them.

timetables rosters workshops

See the status →

What about your company?

If one of these experiments reminds you of a problem of yours: accumulated history nobody uses, decisions that depend on a veteran's memory, data nobody looks at, write to us and we'll talk it through. No fluff: we'll tell you whether it can be done.