Research · 1999 → 2026
Research & Results

Giordano Cabral has researched artificial intelligence since 1999: at CIn/UFPE, at LIP6 of the Université Pierre et Marie Curie (Paris 6), at Sony CSL Paris, at UFRPE and at Stanford University. Since 2025, at FAROL, a research group he founded with Filipe Calegário at CIn/UFPE and which in 2026 becomes the FAROL Institute of Agentic Fluency, he has measured how organizations and people work with AI that carries out whole tasks (it uses tools, chains steps and acts on systems). He chaired the Board of Directors of CESAR from 2020 to 2026 and took part, as a municipal councillor for Science, Technology and Innovation, in validating the guidelines of the Municipal Artificial Intelligence Strategy of Recife (March 16, 2026).
Before FAROL and alongside it, he applied AI to computational creativity and to education, with music as the most studied case: systems that generate music and audio in real time, from the master's degree at UFPE (2002) to the doctorate in generative artificial intelligence at LIP6, with Jean-Pierre Briot and François Pachet (2008); audio analysis, from music information retrieval to the detection of gunshots and explosions for public safety; digital musical instruments; digital games as a measuring instrument in decision experiments; multimedia and mixed reality; and educational technologies.
01 · Research & Results
Results
- 7startups linked to MusTIC and to the group that preceded it, from 2000 to 2019, among them Daccord, Batebit and Audio Alerta (MusTIC presentation at SBCM 2019)
- 80+products launched by these companies
- 116countries with users of Daccord products, more than 4 million people, according to the company (2016)
- 23games released with the audio API built with MusiGames (MusTIC paper)
- 3patents at INPI, one of them with Samsung and UFPE (Google Patents)
- 16theses and dissertations supervised at CIn/UFPE and UFRPE, from 2010 to 2025 (supervisions)
- Flow Machines. Brazyle (2017), a database of Brazilian music, is the Brazilian part of the Flow Machines project, from Sony CSL Paris, on artificial intelligence for music creation, the result of three years of cooperation with the Sony team, Flow Records and later Spotify. The project produced the album Hello World (2018), one of the first records created with artificial intelligence (album credits).
- Daccord. D'Accord Guitar, the system from the master's dissertation that generates the guitar performance from the chord chart, became the first product of Daccord Music Software and reached 2 million users by 2003 (MusTIC presentation at SBCM 2019). Later came the chord dictionary, the programs for learning guitar, keyboard and drums, and iChords, which recognizes the chords of a recording (iChords on Google Play).
- Public safety. Automatic audio classification, researched in CNPq and FACEPE projects from 2010 to 2014, led to the Audio Alerta systems, which couple microphones to surveillance cameras to detect, classify and locate gunshots and explosions (MusTIC paper, §3.8).
- Games. Drums Challenge HD, from MusiGames, the games business unit, was the best-selling music game for the iPad in the United States and the United Kingdom in 2010 (JC Online, April 4, 2012).
- Schools. Turma do Som, from the Digital Platform for Music Education (CNPq and FINEP, 2010–2014), brought 25 music-education games to schools in Pernambuco, São Paulo and Minas Gerais and was a finalist for the 2015 Jabuti Prize, in the Digital Children's Book category (press kit, 2012; Escribo, 2015).
- Paper awards. Best student technical paper at the 11th SBCM (2007), with HarmIn applied to the work of Tom Jobim, and best paper at the 17th SBCM (2019), with TumTá and Pisada (CIn portal).
- Conferences and journals. IJCAI 2015, in a demonstration with Sony CSL Paris; ISMIR 2005 and 2016, and the invited session Why Is Brazilian Guitar Interesting? at ISMIR 2013; NIME 2013, 2015, 2019 and 2021; SIGGRAPH 2010 (poster); IEEE ISM 2008 and 2009; HCI International 2014. In journals, Organised Sound (2013), IEEE MultiMedia (2017, with McGill and Inria), Computers in Human Behavior (2017) and Behavior Research Methods (2021) (publications).
- Public policy. Lead researcher of Preparing Brazil's Youth, from Stanford Impact Labs (2025), which evaluates vocational education in more than 130 public schools in Pernambuco through a randomized experiment (Stanford Impact Labs).
- FAROL. Open instruments for measuring fluency in agentic AI: the six-axis scale, with a bank of 264 items; the AI Index for companies, with no registration; Meu Mapa de IA, for students; and Arena das IAs, in which 124 AI models predicted the 104 games of the 2026 World Cup, from 06/11 to 07/19, in the same ranking as people (farol-ia.org; Arena das IAs).
02 · Research & Results
Research areas
1. Agentic artificial intelligence and organizations (FAROL) · 2025–. Measuring how organizations and people work with AI that carries out entire tasks; the six-axis scale, the AI Index for companies, the open instruments and the Arena das IAs (2026).
2. Real-time music and audio generation · 1999–2015, resumed in 2021. From the D'Accord Guitar (2001) to the thesis Harmonisation automatique en temps réel (2008) and to Ramin; the application demo at IJCAI 2015 with Sony CSL; Brazyle (2017).
3. AI-based audio analysis: music information retrieval · 2005–2019. The validation of EDS in chord recognition (ISMIR 2005), HarmIn (2007), audio monitoring for public safety at UFRPE and the MIREX evolution index (ISMIR 2016).
4. Digital musical instruments and MusTIC · 2011–. Sketchument, Illusio, the prototyping method and toolkit developed with McGill and Inria (IEEE MultiMedia, 2017), the two VioLED patents, TumTá and Pisada.
5. Games, decision-making and social dilemmas · 2013–. The Decisores Virtuais (Virtual Deciders) project on computational decision models, the PsyFun group and the two games validated as a prisoner's dilemma for children.
6. Multimedia, computer vision, and mixed reality · 2005–. From the framework for percussive systems (2005) to computer vision for diagnosis and for motion analysis, to the mixed reality protocol on laptops with Voxar Labs (2022), and the mixed reality patent with Samsung and UFPE (2024).
7. Education and educational technologies · 2008–. Turma do Som, the Ática book, games and virtual reality at school, gender-sensitive computational thinking, the period as a Visiting Scholar at Stanford and the survey of how students work with AI.
03 · Research & Results
The groups
- FAROL: research group founded in 2025 with Filipe Calegário, at CIn/UFPE, which in 2026 becomes the Instituto FAROL de Fluência Agêntica (FAROL Institute of Agentic Fluency). It measures the fluency of organizations and people in agentic AI. farol-ia.org
- MusTIC: Music, Technology, Interactivity and Creativity, a CIn/UFPE research group created in 2011, coordinated by him with Geber Ramalho and Filipe Calegário. mustic.cin.ufpe.br
- PsyFun: digital games for cognitive psychology. Started by him at UFRPE, in October 2013; today it sits between CIn/UFPE and UNIVASF, with Guilherme Ribeiro Eulálio Cabral.
- Voxar Labs: virtual and augmented reality laboratory at CIn/UFPE, directed by Veronica Teichrieb, with whom he collaborates. voxarlabs.cin.ufpe.br
04 · Research & Results
Collaboration and service
Cooperation with France has been continuous since 2001: the Smart-es project (UFPE × Université Paris VI, 2001–2004, led by Geber Ramalho), the PhD in generative artificial intelligence at LIP6, with Jean-Pierre Briot and François Pachet (2003–2008), the thesis committees at IMPA with both of them (2009 and 2010), Briot on the committee of Eduardo Santos Silva (2022), and the three texts on AI in Brazil deposited in HAL in 2024. With the United States: Visiting Scholar at the Lemann Center, at Stanford University, from September 2023 to August 2024, and lead researcher of the project Preparing Brazil's Youth, of Stanford Impact Labs (2025). With Canada and France, the co-authorships with Marcelo Wanderley (McGill) and Stéphane Huot (Inria) between 2015 and 2017. He was an Advisory Board member (Advisory Board) of the European project PRAISE (FP7, 2012–2014) and a project reviewer for CNPq in 2011, 2012 and 2013.
In the Brazilian Computer Society: SBCM technical committee in 2007, 2009 and 2011, technical committee chair in 2013, local organizer of the 12th SBCM (Recife, 2009) and general coordinator of the 18th (2021, with Filipe Calegário); steering committee of the Special Committee on Computer Music from 2011 to 2014 and from 2021 to 2022, and chair of the Committee in 2022–2023; technical committee of ICMC 2010.
05 · Research & Results
By the numbers
Counts from Lattes (08/19/2026), the CIn portal, and citation databases (09/24/2026). The bibliographic output begins in 2001: there are 88 unique works through 2023; those from 2024 to 2026 are in Publications.
| Journal articles | 10 on Lattes; 11 in the databases |
|---|---|
| Full papers in conference proceedings | 56, plus 2 extended abstracts |
| Book chapters | 2 |
| Books | 2 authored (Ática, 2014; Escribo, 2015, with Américo Amorim); 28 records of edited books (7 textbook titles and 2 proceedings); 1 preface |
| Computer programs | 93 |
| Patent applications filed | 3 at INPI: two for VioLED (2017) and one for mixed reality with Samsung and UFPE (2024) |
| Best paper awards | 2 (SBCM 2007 and 2019) |
| Completed supervision | 2 doctoral and 14 master's students as primary advisor; 2 doctoral and 10 master's co-advising records; 8 undergraduate final projects; 10 undergraduate research projects; 1 postdoctoral researcher |
| Examination committees | 25 master's, 14 doctoral, 15 undergraduate, 3 faculty recruitment committees |
| Recorded projects | 8 research, 14 technological development, 7 outreach and teaching (CNPq, FACEPE, FINEP, BNB) |
| Citations | 268 on Semantic Scholar (h = 10); 194 on OpenAlex (h = 8, i10 = 5) |
06 · Research & Results
Images from the archive


AdvisingPhD, master's, and undergraduate · 2010 → 2025
Two theses and fourteen dissertations as principal advisor; one PhD and ten dissertations as co-advisor; eight undergraduate capstone projects, ten undergraduate research projects, and one postdoctoral fellowship.
Frequently asked questions
What does Giordano Cabral research?
Artificial intelligence: agentic AI and its adoption by organizations, at FAROL (2025–), and the application of AI to computational creativity (real-time music and audio generation, from 1999 to 2015; automatic audio and music analysis; digital instruments; games and multimedia) and to education. The groups are FAROL, MusTIC, PsyFun and Voxar Labs.
What results has the research produced?
Seven startups linked to MusTIC and the group that preceded it, from 2000 to 2019, with more than eighty products, those from Daccord used by more than 4 million people in 116 countries, according to the company; 23 games released with the audio API built with MusiGames; three patents, one of them with Samsung and UFPE; Brazyle, the Brazilian part of the Flow Machines project at Sony CSL Paris; two best paper awards at SBCM (2007 and 2019); sixteen theses and dissertations supervised; and, at FAROL, open instruments for measuring fluency in agentic AI and the Arena das IAs (2026).
What is FAROL?
A research group on agentic artificial intelligence founded in 2025 by Giordano Cabral and Filipe Calegário at the Informatics Center of UFPE, which in 2026 becomes the FAROL Institute of Agentic Fluency. It measures how organizations and people work with agentic AI and publishes open instruments, studies and training at farol-ia.org.
What does FAROL measure?
Fluency in the adoption of agentic AI by companies, schools, and government, across six axes. The instrument measures fluency, understood as movement, rather than maturity, understood as position; the results are used in diagnostics, training, and open instruments.
What is MusTIC?
A research and innovation group at CIn/UFPE focused on music, technology, interactivity and creativity, co-led with Geber Ramalho and Filipe Calegário. Its work includes TumTá and Pisada (best paper at the 17th SBCM, 2019), Batebit Controller, Sketchument and Probatio.
What is PsyFun?
A group that builds short games to measure decisions in social dilemmas (cooperation, common resources, game theory). Started at UFRPE in 2013, now between CIn/UFPE and UNIVASF, with results in Computers in Human Behavior (2017) and Behavior Research Methods (2021).