David Davó Laviña

Under construction!
Information in this website may be incomplete or untranslated

David Davó Laviña

Research Assistant at GRASIA
Author Image

Research assistant that just finished its Master’s in Artificial Intelligence at the Politechnic University of Madrid. Currently I am searching for opportunities to finance a PhD.

While I studied things more related to theoretical computer science and artificial intelligence, I also like hardware and computer architecture. I am a software libre, hacking and computer history enthusiast. My previous publications are related to Human-Computer Interaction, but I am eager to learn new things.

In my free time, I have a small homelab and study Japanese, and I also have a blog.

Projects

InvProy α
High School Student 2015 - 2017

Final project part of the “Baccalaureate of Excellence” program, winner of the 2nd pre-university meeting of future researchers (2017).

SIMKL's TV Tracker
3rd Party Developer 2016 - 2019

Creation of plugins for Emby and Kodi that scrobble the chapter or movie you’re watching to Simkl

Just Get My Data
Grad Student 2022 - Today

Repository of links and documentation on how to obtain your personal data from multiple web services

IHaskell-Widgets
Contributor Verano 2021

Fixing the IHaskell-widgets library as a participant in the Google Summer of Code 2021 with Haskell.org as mentor

wdsparqlmagic
Developer 2022

Jupyter plugin to execute SparQL queries inside the notebook

Bellman Ford in Haskell
Grad Student 2020

Haskell implementation of the Bellman-Ford algorithm, analysing its complexity and efficiency

Artificial Intelligence Notebooks
Grad Student 2019

Jupyter Notebooks of Artificial Intelligence from my BSc in Computer Science

DAO-Analyzer
Developer 2021 - 2022

Dashboard to display data on Ethereum’s Decentralized Autonomous Organizations (DAOs)

Telephone Tales & Poems
Organizer & Developer 2015 - Today

Website to manage requests in the “Telefone Tales” event

Homelab
Sysadmin 2011 - Today

Small home network with multiple servers that hosts various daily use services for friends and family, using docker and lots of scripting and duct-tape

TFG: NoC implementation of a RISC-V processor
Hardware design

Implementation of a Network on Chip for the SWerv-EL2 processor written in Verilog

Publications

Exploring recommender systems for proposal recommendation in Decentralized Autonomous Organizations

DAOs have emerged as a novel approach to collective governance, facilitated by blockchain technology. However, the effectiveness and legitimacy of decision-making within DAOs can be compromised by low voter turnout. Given the limited lifespan of each proposal, conventional recommender system techniques are unsuitable. In response to these issues, this Master’s Thesis introduces a recommender system specifically designed for DAOs.

Decentralized Autonomous Organizations (DAOs) are a new kind of organization that relies on blockchain software to govern their projects. Typically, DAO members may put forward and vote on proposals. For instance these proposals may consist on someone doing some tasks in exchange for a share of the DAO crypto-funds. In recent times, DAOs have gained a remarkable adoption, and yet they are still understudied by the academic literature. In this work, we present a visual analytics tool to study DAO activity focusing on their participation and temporal evolution. Our tool will hopefully help to stimulate research on this new kind of online community and collaborative software.

Network on a Chip (NoC) is an emergent technology alternative to traditional inter connection methods, in which network design methods are applied to the connexions between modules. In this Final Degree Project, we apply this technology inside the execution unit of the RISC-V based processor SWerv-EL2. For this, we have studied in-depth and designed a NoC, trying to minimize the resources utilized and its impact on the core. Afterwards, we modified the execution unit of the core to include said network, using it to interconnect its submodules. In both cases, we used the hardware description language SystemVerilog.