Datox

Login Page
Dashboard Stats Page

Inspiration

Countless times, we have realized just how helpful AI could be to our projects. Yet more often than not, we have to drop our plans because they would require specific data that seemed impossible to find online. For instance, at one point I was in great need of segmentation data on old geographic maps. Or another time, we had to abandon an otherwise great project just because of a lack of data regarding financial advice. Due to these frustrating concerns, we have made a platform which streamlines the data acquisition process like never before.

What it does

Datox allows customers to get their data needs fulfilled by countless freelancers on the site. For most complex data needs, this platform would be perfect as it reduces the complexity of this task by tenfold for the customers.

How we built it

We use Next.js and Tailwind to produce the main website of this platform. As Data is at the very core of Datox, we needed to ensure unshakable database integrity. With this in mind, we settled upon PostgreSQL. Furthermore, aside from the main website, we employed a Python API to do most of the complex file handling tasks.

Challenges we ran into

The data management in general was a complex task. Since it was such a core functionality, we simply could not have created a half-baked feature. The reason it was challenging was in no small part because of differences between production and deployment environments. Due to this, deploying our platform took unexpectedly long. Though we have experience deploying before, the considerable storage requirements of this app made this challenging on free-tier hosting providers.

Accomplishments that we're proud of

Deciding to use PostgreSQL was an improvised decision, yet we managed to learn it sufficiently well. Furthermore, we are proud that we were able to get the first version of our app fully made and deployed, as often time constraints can force an incomplete product. Finally, we are happy with the execution of our front-end.

What we learned

We have definitely learned to work with databases and file systems much better as we needed to ensure that this system is robust. Furthermore, there were many different components to this platform, and we had to distribute the work amongst ourselves. However when merging all of our code, we realized the value of being in sync. In addition, we will certainly keep in mind for future reference how unpredictable deployments can be, as that took up a rather high percentage of our time.

Market Research

We wanted to conduct some Market Research to see how our project might fare in the real world. We got in touch with representatives from Epic Systems and Cobalt Robotics and asked for their opinion on the viability and usefulness of this project. They both expressed interest in the idea. The representative from Cobalt stated that Datox could resolve problems by taking the burden of labelling off of engineers and could be popular with freelancers looking for a "gig" job. The representative from Epic Systems said that if challenges like labelling verification could be overcome, the platform could improve efficiency with its centralization.

What's next for Datox

The current era witnesses a rapid expansion in the AI field, and in this context software such as Datox will be crucial in facilitating the pursuit. Therefore, there is definitely a large opportunity for Datox, and upon doing some research, this has become ever more clear. According to this study , about 96 % of enterprises encounter data quality and labelling challenges. Furthermore, it even states that teams that outsource data labeling get into production faster. Therefore, it is clear that the industry needs more effective solutions to tackle this challenge, and Datox will do just that by making it easier than ever to outsource labeling and maintain its quality. By enabling customizable payouts, Datox operates on Laissez-faire economics, converging towards market equilibrium, ensuring competitiveness and attractiveness for both freelancers and customers. Despite taking a small transaction cut for profitability, Datox will remain stable due to projected growth in the freelance market (≈16%) which will offset for the marginal dip in supply of freelancers caused by the transaction cut. Consequently, Datox appears poised for success as a stable, thriving, and profitable product.