About
I’m Daniel Bastos Pereira — a software developer since 2003 and, these days, a Senior Technical Consultant at VILT in São Paulo, Brazil. I’ve built desktop, web and mobile software, and somewhere along the way I started experimenting with Artificial Intelligence with real-world use cases. This blog — Hikari Tec BR — is my own corner of the web.
What I’m working on now
Most of my attention right now goes to AI-assisted content production on the Adobe Experience Cloud. I work on the pre-sales and solution-architecture side of it: helping enterprise teams figure out where generative AI and agentic workflows actually fit their content supply chain, scoping the work, and building proof-of-concepts before anyone commits to a roadmap.
Two threads run through most of that work. The first is agent development — designing and building the agents that automate work pipelines, increasingly with frameworks like Google’s Agent Development Kit (ADK). The second is GEO — generative engine optimization — the technique for producing content that AI answer engines will actually surface.
Underneath it all I’m a firm believer in spec-driven development: get the specification right and the implementation — human or agent — will be solid.
How I got here
A quick tour, newest first:
- 2020–present — Senior Technical Consultant, VILT. Consulting, tech lead and development on web projects across the Adobe and OpenText stacks — pre-sales, solution architecture, PoCs, web analytics implementations, and in-company training.
- 2023–2024 — Senior Support Engineer, AmeXio (part-time, remote, alongside VILT). AEM support for a Belgian media group’s investment and content websites.
- 2018–2020 — Senior Software Engineer, Pipa Studios. Java backends for mobile games across iOS, Android and Facebook, running on AWS.
- 2016–2018 — Senior Technical Consultant, VILT. My first tour — AEM and OpenText WEM development and solution architecture.
- 2014–2016 — Software Engineer, Élin Duxus Consulting. Software for risk analysis and credit rating.
- 2009–2014 — Co-Founder, Mindexs Mobile. Built and ran a small mobile-software shop — the code, the hiring, and the business plan.
- 2008–2013 — Itaú Unibanco. Channel analytics for ATM, internet banking, mobile and IVR — from intern to business analyst, on projects like the Itaú/Unibanco merger and the biometric-ATM rollout.
- 2006–2009 — Research scholarship, Universidade de São Paulo. Natural Language Processing research, with a few published articles — my first run at getting machines to deal with human language. Funny how that one keeps coming back around.
Academic work
Before consulting, I did research. From 2006 to 2009 I held a scholarship in the natural language processing group at the University of São Paulo (USP) — advised by Ivandré Paraboni — while earning my bachelor’s degree in Information Technology there. The work was on statistical natural language generation: getting machines to turn structured data into fluent, human-readable text.
It collaborated with these publications:
- 2010 — D. J. Lucena, D. B. Pereira, I. Paraboni. “From Semantic Properties to Surface Text: the Generation of Domain Object Descriptions.” Inteligencia Artificial, vol. 14, pp. 48–58.
- 2010 — R. P. A. Araujo, R. L. Oliveira, E. M. Novais, T. D. Tadeu, D. B. Pereira, I. Paraboni. “SINotas: the Evaluation of a NLG Application.” 7th International Conference on Language Resources and Evaluation (LREC 2010), Valletta, Malta.
- 2009 — E. M. Novais, R. L. Oliveira, D. B. Pereira, T. D. Tadeu, I. Paraboni. “A Testbed for Portuguese Natural Language Generation.” 7th Brazilian Symposium in Information and Human Language Technology (STIL 2009), São Carlos, Brazil.
- 2008 — D. B. Pereira, I. Paraboni. “Statistical Surface Realisation of Portuguese Referring Expressions.” GoTAL 2008 — 6th International Conference on Natural Language Processing, Gothenburg, Sweden. Springer LNCS, vol. 5221, pp. 383–392.
- 2008 — F. M. V. dos Santos, D. B. Pereira, I. Paraboni. “Rule-based vs. Probabilistic Surface Realisation of Definitive Descriptions.” VI Workshop on Information and Human Language Technology (TIL), Vila Velha, Brazil.
- 2008 — D. B. Pereira, I. Paraboni. “From TUNA Attribute Sets to Portuguese Text: a First Report.” 5th International Natural Language Generation Conference (INLG 2008), Salt Fork, Ohio, USA.
- 2007 — D. B. Pereira, I. Paraboni. “A Language Modelling Tool for Statistical NLP.” 5º Workshop em Tecnologia da Informação e da Linguagem Humana (TIL), XXVII Congresso da Sociedade Brasileira de Computação, Rio de Janeiro, Brazil.
My full academic record — including conference presentations and research software — lives on my Lattes CV.
Elsewhere
- GitHub —
@bastdan - X / Twitter —
@bastdan - Lattes — my academic CV and NLP research record