Table Colombia Challenge

• English
• Spanish

The best way to build performant AI is by creating qualified training data. The need for tools and infrastructure to ensure that the right information is used to train these AI systems is unavoidable. Building AI applications is hard and expensive, and largely because most AI teams today are spending 80% of their time building tools and infrastructure to create and manage training data.

Pain pints:

Tools : What is already out there is quite expensive and requires specific knowledge and training.
Workforce : Managing your own team and training them on a variety of tagging requirements is time consuming and expensive.
Quality : You need to develop a QA model to make sure your algorithm is learning the right way.

LinkedAI is a collaborative annotation platform powered with AI, bringing data annotation and data management in one single place. Our labeling platform trained with AI reduce annotation time in more than a 50% to focus in the most important part of creating training data: quality.

LinkedAI also provides dedicated labeling services, we achieve best-in-class data quality by auditing worker annotations against tasks with known answers and triple checking the annotation quality. In contrast to crowdsourcing platforms, we offer an ethical payout and protection to our workers, who receive a stable monthly salary, an office space and trainings. We provide employment to women who are heads of families and other vulnerable groups.

Artificial Intelligence market is fast-expanding valued in more than 10B USD with a CAGR of 48% between 2018 and 2023.

Tractica forecasts that the revenue generated from the direct and indirect application of AI software is estimated to grow from $643.7 million in 2016 to $36.8 billion by 2025. This represents a significant growth curve for the 9-year period with a compound annual growth rate (CAGR) of 56.8%.

Classification and image tagging or labeling is in second place of Tractica’s top 10 AI use cases in terms of revenue in 2025, 40% are related to image or object recognition, which is a good thing for LinkedAI, as we make part of this use cases.

We don't have competitors in Colombia or Latin America for the moment. Our principal competitors are based in the US, Scale ai, Label Box, Samasource and Amazon Mechanical Turks, are some of our strongest competitors.

Our AI powered platform recognizes a great variety of objects within images to reduce annotation time, we can automate a great part of the labeling process using AI algorithms, in contrast to some of our competitors.

We can increase the accuracy of AI models by automatically  generating valuable synthetic data to enhance datasets, a great competitive advantage to our competitors who can't assure accuracy increase.

We combine proprietary tools with our workforce to provide ground-truth data every time.

In contrast to crowdsourcing platforms, we offer an ethical payout and protection to our workers. We provide employment to women who are heads of families and other vulnerable groups.

We have 110 registered users in our platform and at least 20 of them upload and label their datasets daily. Our platform is free for the moment, this allows us to capture labeled data from our users to keep training our algorithms. We generate revenue by providing dedicated labeling services, we have worked with companies from the US and Colombia, Asimetrix Inc, River Co, Coca-Cola, and Transport Systems are some of them.
- Users from +11 different countries, most actively in India and USA.

We generate content for people and companies working with AI and Computer Vision, we share this content through our social media and different platforms as Medium and Quora, this have allowed us to get to both users and customers. We are proud to say that our last customers came directly to us and we haven't spent as dollar in Ads. We also use emailing and LinkedIn strategies by targeting companies working with AI and Computer Vision as their core product.

We are closing a partnership soon with a Canadian/Indian company which provides labeling services to companies around the world, for them to use our platform and they will support us with their workforce in case we need it.

Our principal KPI is # of labels made in our platform

- 7x growth in June and 9x growth in July (strongest months in growth rate)
- +200.000 labels done in the last three months.

44.5% Paula Villamarín (CEO) Product designer with emphasis in tech products. Entrepreneurship studies at UC Berkeley Skydeck, passionate about tech applications, Artificial Intelligence and user experience design. Web programmer, and City AI ambassador. 

44.5%Diego Parra (CTO) Fullstack Developer. Extensive experience developing Apps and web apps in multiples companies like Globant, Colpatria, Urpin, Higuera Studios for the last 10 years. Fluent in multiples programming languages including java, c++, js and more. Active member of City AI. Extremely passionate about programming, open source, video games, life-long learning.

1% Cristian García (ML Advisor) is a Machine Learning Engineer and Developer with background in math and physics. Extensive experience creating Deep Learning applications in fields such as autonomous vehicles, video analytics and manufacturing. He's expert on several frameworks including Tensorflow, Numpy, Keras, Scikitlearn and Spark. Community leader and active speaker at conferences world-wide, host of the Machine Learning Meetup Medellin, and admin of Machine Learning Colombia. Extremely passionate about programming, open source, deep learning, life-long learning. Member of Toptal: top 3% of the developer talent in the world.

Second semester 2019- Capital needs 42k, revenue 12k, 4 paid clients and 25 active users (free version) in the platform. Milestone: 3'000.000 labels done in our platform.

2020- Capital needs 250k, revenue 60k, 8 paid clients and  80 active users (free version) in the platform. Milestone: 10'000.000 labels done in our platform.

First semester 2021- Capital needs 200k, revenue 100k, 12 paid clients and  160 active users (free version) in the platform. Milestone: 22'000.000 labels done in our platform.

Yes! LinkedAI has an impact in Colombia, where we employ women who are heads of families and other vulnerable groups, to make part of our labeling team. In contrast to crowdsourcing platforms, we offer an ethical payout and protection to our workers, who receive a stable monthly salary, an office space and trainings. We sure want to continue helping vulnerable groups in Colombia, that's why we are looking to partner with non-profit foundations, jails and projects in the region to continue expanding and generating more impact everyday.