SpainHealthcare and life sciences2026 – ongoing

PharmaMar

Multi-agent platform to prioritise oncology drug candidates

PharmaMar and Globant built a 20-agent AI system that picked 10 drug-indication combinations from more than 8,000

In progressVendor case studyVendor ↗

Problem

PharmaMar's researchers had to weigh internal data, scientific publications and regulatory sources to choose which cancer drug candidates to pursue. The volume of material made that slow.

Approach

PharmaMar worked with Globant to build a multi-agent system on Globant Enterprise AI. More than 20 specialised agents review internal databases, publications and FDA and EMA sources. A next phase is meant to add hypothesis generation, regulatory compliance checks and automated report drafting, with no timeline given.

Results

More than 4,500research documents reviewed by the system, per Globant and PharmaMarVendor case study[1]

10 of more than 8,000treatment-indication combinations prioritised from the possibilities reviewed, per Globant and PharmaMarVendor case study[1]

Up to 15 timesreduction in time to obtain key insights, as claimed in the releaseVendor case study[1]

More than 90%accuracy in retrieving complex data, as claimed in the releaseVendor case study[1]

Takeaway

The speed and accuracy figures come from the vendor and the company. No drug outcome has been reported yet.

Sources

PharmaMar busca acelerar la investigación oncológica con IA a través de una colaboración con Globant
PR Newswire19 Mar 2026Vendor publication

Source summaries

[1]PR Newswire19 Mar 2026

The release says Globant and PharmaMar created a multi-agent system on Globant Enterprise AI for oncology research. It cites more than 20 agents, 4,500 documents reviewed, 10 combinations chosen from over 8,000, up to 15 times faster insights and over 90% retrieval accuracy. It also lists planned next-phase features.

Full details are in the original sources.

Added 1 Nov 2026 · Updated 1 Nov 2026