Website: [https://cleca.app](https://cleca.app)

First line of code: February 2022

Made the easiest tool to help my girlfriend learn Catalan. Because all language learning apps to learn catalan suck and have too much friction, I created a WhatsApp bot that every day it sends a new word, expression together with translation and pronunciation.

![Screenshot 2022-07-29 at 17.19.52.png](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/Screenshot_2022_07_29_at_17_19_52_5c3debd5c4.png)

#TailwindCSS #NextJS #SoftCatala #GCP

A WhatsApp bot that every day it sends a new word, expression together with translation and pronunciation.

Cleca

**Hackathon project**

Open-source: [Github repo](https://github.com/jamalavedra/encode-podquest)

Project presentation: [View the final pitch on youtube](https://www.youtube.com/watch?v=XojNGTNY-h4)

For the [Encode](https://www.encode.club/) Polygon hackathon, I created a Mirror for Podcasts that uses Lens protocol. The podcast itself is stored in IPFS and published as a post (so other platforms can choose to display it) and includes crowdsourcing functionalities for the creator

It includes social features like comments and likes that would help people find out the most interesting content on the platform.

One of the most interesting value props of web3 is incentive alignment, so I decided to create a crowdsource feature so that creators can ask interested people to stake their interest in giving them money, Creators can also monetize by allowing viewers to mint an NFT from the podcast. One of the main goals that I was excited to explore was, is there a way to set bounties like what [dework](https://dework.xyz/) does for tasks but to incentivize podcasts being created around topics we find interesting. For example:
- bounties so that a podcast is done around a topic we like.
- bounties so that a certain person does voice over a certain podcast
- bounties so that a podcast gets translated.

[Encode x Polygon Hackathon: Finale Prizewinners and Summary](https://medium.com/encode-club/encode-x-polygon-hackathon-finale-prizewinners-and-summary-482a19848535)

Hackathon project: Mirror for Podcasts that uses Lens protocol. The podcast itself is stored in IPFS and published as a post on Lens

Podquest

**Hackathon project**

Open-source: [Github repo](https://github.com/jamalavedra/rward-SDK)

For the [NYC Ethglobal](https://ethglobal.com/) hackathon, with [Joan](https://twitter.com/joalavedra), we implemented an SDK to help developers to quickly and effectively integrate micro incentives in their projects.

Gamification has been the backbone of any user experience in software. Games, SaaS, consumer apps, etc. all combine design mechanics and incentives to accomplish objectives and goals. We believe that crypto enables new collaboration games where users can also be rewarded with tokens. 

Tokens represent a paradigm shift in the way we build communities, software and ownership. Enter Rward a simple API that enables developers to integrate backed tokenomics without the complexity of crypto. 

Rward handles the sending tokens with rward.rewards.send(address, amount) and wallet creation with rward.customer.create(email). It abstracts away complexities for users that don't know about web3 with a non-custodial wallet powered by Privy. 

A wallet is automatically created on their behalf of the user and they can claim it with their usual OAuth.

The project uses several technologies to make this SDK possible. Tatum is used to generate the wallet and to send the ERC20 tokens. Privy is used to store the seed phrase, the private key, email and address to make sure that only the user that owns a certain email can claim a Wallet.


[EthGlobal showcase](https://ethglobal.com/showcase/rward-wfpza)

Hackathon project: Implemented an SDK to help developers to quickly and effectively integrate micro incentives in their projects.

Rward

**Hackathon project**

Project presentation: [View the final pitch on youtube](https://www.youtube.com/watch?v=5qD-Vdu7hsM)

For the [Encode](https://www.encode.club/) DAO hackathon, with [Joan](https://twitter.com/joalavedra), we created a way for DAOs to organize and rank information that is shared from within the organization.

The main idea was to make content reactive to a search on a search engine by displaying the matching links that contained relevant information. Communities then could vote, a-la Reddit, what content is more relevant for the organization, rank it and comment on it. Everything happening on-chain thanks to Lens protocol

[Encode x BitDAO: DAO Hack Finale Prizewinners and Summary](https://medium.com/encode-club/encode-x-bitdao-dao-hack-finale-prizewinners-and-summary-825d7fb11c30)

Hackathon project: A better way for DAOs to organize and rank information that is shared from within the organization.

Onemark

**Hackathon project**

Open-source: [Github repo](https://github.com/jamalavedra/ethglobal-LFGrow-aura)

Project presentation: [View the final pitch on youtube](https://www.youtube.com/watch?v=GaWE33EJFBg&t=3947s)

For the [LFG Ethglobal](https://ethglobal.com/) hackathon, with [Joan](https://twitter.com/joalavedra), we implemented a reputation system for Lens protocol. 

It creates a Reference Module that checks the on-chain reputation of a certain user based on its activity. When a user initializes the module, they can choose the threshold they are willing to accept for comments to be made to their post. It that score is fulfilled when trying to comment, that is allowed. 

Reputation is computed off-chain and accessed via a Chainlink any API. It is contemplated that other lens information will be used to compute the reputation of a certain user such as sentiment analysis of their older comments and posts.

This project creates a new reference module for lens protocol. It uses chainlink any API to query an API endpoint that gathers on chain information about a certain address. We use The Graph for indexing and querying this information. 

It only allows a user to comment on a post given that they have enough reputation based on what the author of the post has decided to be the threshold

[Messari overview of Lens](https://messari.io/article/the-open-social-map)

[EthGlobal showcase](https://ethglobal.com/showcase/aura-reputation-unp5v)

Hackathon project: Implemented the first reputation system for Lens protocol. 

Aura Reputation

First line of code: August 2021

With [Joan](https://twitter.com/joalavedra), we created a tool to help companies place backlinks in relevant websites and therefore rank higher in Google. Companies could identify relevant opportunities by entering keywords that are relevant to they business, and we would analyze journalists requests looking for sources and score them on the keywords. 

![Screenshot 2022-07-29 at 17.17.38.png](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/Screenshot_2022_07_29_at_17_17_38_49a9b78cd8.png)

It works by streamlining the discoverability of reporters searching for a source and helping SME get in touch with them. With the help of OpenAI's GPT-3 models, we would write compelling pitches to the journalists so that companies could get in touch with more outlets.
![Screenshot 2022-07-30 at 13.37.04.png](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/Screenshot_2022_07_30_at_13_37_04_18777ed92f.png)

We grew this project through outreach on LinkedIn and creating tools that would help companies evaluate their opportunities in SEO.
![Screenshot 2022-07-29 at 17.14.11.png](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/Screenshot_2022_07_29_at_17_14_11_1bc6ad6903.png)

We sold HeadcomAI in February 2022.



#TailwindCSS #OpenAI #Firebase #GCP #NextJS

A SaaS platform to help companies improve their SEO by getting link placements in journalists' articles that are looking for sources

HeadcomAI

Link: [https://onsite.fun](https://onsite.fun)

First line of code: October 2020

With [Joan](https://twitter.com/joalavedra), we created Onsite.fun to help remote teams create and mantain camaraderie when distributed.

We developed this B2B marketplace from the ground up to accomodate the needs of companies when booking online events for their remote teams. We implemented some cool features like budget control, groups and memberships. Check out a full list of features [here](https://onsite.fun/features).

![Screenshot 2022-07-29 at 17.18.49.png](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/Screenshot_2022_07_29_at_17_18_49_d8530bfe34.png)

During this time we also experimented in releasing a breathwork app under the Onsite.fun umbrella. You can check it out at [Spiro](https://www.joinspiro.com/).





A marketplace of online team-building events for remote companies to create camaraderie

Onsite.fun

First line of code: June 2020

Peerpull is a platform for entrepreneurs and digital makers to connect. You can connect with people that have gone through similar doubts or issues by simply creating a room and joining at the scheduled time. You can also interact with others by joining the rooms other peers create.

Here's a GIF I managed to keep from one of the earlier versions:
![peerpull.gif](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/peerpull_ed8a4635fc.gif)

Almost at the end of this project, we integrated Algolia to help with the discovereing of relevant people that could help in solving your doubts:
![Screenshot 2022-07-30 at 13.40.09.png](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/Screenshot_2022_07_30_at_13_40_09_d740a7ba62.png)

#ReactJS #GraphQL #AWS #S3 #GCP #JitsiMeet #Algolia

Connect with people that have gone through similar doubts or issues by simply creating a room and joining a call at the scheduled time. 

PeerPull

## Development of feature based smart virtual sensors

Open-source: [Github repo](http://bitly.ws/cpnV)

This thesis proposes an application of convolutional feature points as smart virtual sensors. The feature descriptors are formed using the feature maps of one of the layers of a VGG-16 network pre-trained on ImageNet.From the extracted and matched features, information is obtained from each of them to create a data-set that describes the sequence of frames under consideration.

#Python #PyTorch

Using convolutional feature points from a deep neural network as smart virtual sensors to measure displacement.

Master Thesis

## Development of an exploration system for mobile robots

Open-source: [Github repo](https://bit.ly/3b9RLfN)

Thesis: [Final hand-in (in Catalan)](https://www.slideshare.net/JaumeAlavedraMas/treball-de-fidegraujaumealavedramas)

During my bachelor thesis, I developed a computer vision-based system in which, through a camera attached to a mobile robot that focuses on the roof, the system is able to describe the movement and relative position of the device.

#MATLAB #ROS

Developed a computer vision-based system in which the system is able to describe the movement and relative position of the device by a camera footage.

Bachelor Thesis

## Formula Student 2016-2017


During this season, I was in charge of the design and implementation of the dashboard and lightning system of the new single-seater for the 2016-2017 season.

The original design of the dashboard I was starting off looked like this:
![original_design_dashboard.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/original_design_dashboard_0f23f5940c.jpeg)

Check out the car from the 2015-2016 season [here](https://www.e-techracing.es/en/ev-a/)

This design had a bunch of flaws. Namely, the harness was very unreliable given that the casing that protected it from rain was subdivided in 3 different boxes that were interconnected.

For the new design, I started off by choosing and selecting a matrix display where it would be possible to show more information and that would also be visible with direct sunlight.

I tested a bunch of displays before making the final decision:
![testing other displays.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/testing_other_displays_6aa4fff5ba.jpeg)

After choosing one, I had to implement the library to control it. I programmed the embeded system in C, and had to implement a library to make it easier to display pixels with it.
![Screenshot 2022-07-30 at 11.49.00.png](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/Screenshot_2022_07_30_at_11_49_00_a0c1e8229a.png)


After some weeks of work, I managed to get it working:
![dashboard_screen.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/dashboard_screen_3811345543.jpeg)


That was the start of the PCB design process, where I was deciding what would be the components that would be included in the dashboard. Finally, I decided for a custom made array of LEDs to indicate battery level, as well as the compliant LEDs for error states and the stop switches. I also added some menu control switches in case the switches we included in the steering wheel would fail.

![design_pcb.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/design_pcb_017d74cb58.jpeg)


The design and printing of the PCBs was pretty smooth, but short after I figured out they had some flaws that I had to mend manually with some nasty soldering on top.

![pcbs_etech.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/pcbs_etech_dbde74dbc5.jpeg)


Fixing issues in the PCB

![fixing issues in the pcb.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/fixing_issues_in_the_pcb_b494ee5abe.jpeg)

After the design of the PCB was finalised, I helped design the enclosing for it. It needed to be dust and water resitant and as easy to manufacture as possible.

![case_design.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/case_design_8f3b32a268.jpeg)


3D render of position of dashboard in car

![3d_render.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/3d_render_a7044a289a.jpeg)


Final fit in the car

![final fit in the car.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/final_fit_in_the_car_1012c5a2c2.jpeg)


Final fit with the wheel

![attached_with_wheel.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/attached_with_wheel_0cc2d447c7.jpeg)


During the competition

![during the competition.jpeg](https://jamalavedra-blog.s3.eu-central-1.amazonaws.com/during_the_competition_c7cad53732.jpeg)


Check out some nice pictures of the final result in at the [eTech Racing website](https://www.e-techracing.es/en/stev-e-2016-17/)

Taking part in the design and construction of a Formula Student electrical single-seater during the 2016-2017 season

🛠 Projects

Podquest

Rward

Onemark

Aura Reputation

Cleca

HeadcomAI

Onsite.fun

PeerPull

Master Thesis

Bachelor Thesis

eTech Racing

Cleca

Hey, I'm Jaume 👋