Methodology
What's our technical step-by-step?
Here's our main flow when it comes to iterating and improving our software engine.
Our Thesis
We’re building a tool to fix any problem that technology can solve - just by talking about it. Adjacent tools exist, though they require immense amounts of expertise and expenses. This includes diverse groups such as youth, who can use it to connect and strengthen their communities, and elderly individuals, who can discover ways to connect and share their life stories.
Validating Foundational Actions
We rely on AI to speed up development by handling routine tasks like testing and debugging, freeing up developers to focus on more complex problems. This is a new sector, with most research applicable to our case being less than 3 years old. McKinsey and Company does a good job at truncating the core benefits with AI in our research.
- AI is proven to accelerate software and ideation development. (https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/tech-forward/scaling-ai-for-success-four-technical-enablers-for-sustained-impact)
- AI sidekicks, or 'Digital Twins,' as McKinsey coined, can be proven to validate the level of difficulty in technical and ideation development and reduce the friction between the idea and bringing it to reality. (https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/tech-forward/digital-twins-when-and-why-to-use-one)
Average Cost
While our goal is offering the lowest possible cost and ease for everyone to solve problems, it comes at a cost. Research and development when it comes to tech has a very high ticket cost. This realization was in fact the reason why Ozone exists now. The only way to know if a software can help pursue our goal is by actually hunkering down and purchasing, developing, then integrating into our logistics flow. No current tool was built with our idea in mind, as they all were built for a world where a heavy knowledge in data engineering and logistics management is needed. This softwares, for our enterprise scale, can cost up to 5-25k per tech stack, and after the cost and hours spent on implementation, we have to cross our fingers to see if it answers one question: Does this make building solutions easier and cheaper for everyone? If the answer is yes, it’s getting implemented! If the answer is no, it’s back to the drawing board.
This is completely uncharted territory, which means that every failure leads me to one step closer to a successful hit. We are currently the only company building this for the average American. For every $20,000 I have placed in R&D, it’s made the end user’s cost 20% cheaper.
Challenges
One of the biggest challenges we face is blending advanced machine learning with user-friendly interfaces that don’t require any technical knowledge. We need to build a system that understands what people say in plain language and turns those words into direct development outlines.
The current tools out there aren’t built with our vision in mind. They’re designed for people who already have a deep understanding of data engineering and logistics management, and they can be very expensive—anywhere from $50,000 to $250,000 per tech stack. Plus, it takes an immense cost of time and money to implement them, and even then, we have to hope they’ll work for our needs.
- Prototyping and Testing: We create and test prototypes of our NLP and AI systems to ensure they can accurately interpret and process user inputs. This involves significant trial and error to refine the technology for seamless integration.
- Integration and Evaluation: We integrate these prototypes into our platform to assess their compatibility and performance. This involves testing whether the software solutions can be effectively read and utilized by our NLPs. If successful, we refine and scale them up. If not, we go back to the drawing board and shelve the software for potential future use.
- A.E.O.S. (Always Embrace Open Source): We leverage open-source technologies to minimize costs and optimize resources, ensuring our platform remains both affordable and capable of integrating various software solutions effectively. If it’s out there and for the public to use, let’s see if the greater public can use it too.
In our case, the only way to know if a software can help pursue our goal is by actually hunkering down and purchasing, developing, then integrating into our logistics flow. No current tool was built with our idea in mind, as the ones needed all were built for a world where a heavy knowledge in data engineering and logistics management is needed.
So far, for every $20,000 we’ve spent on R&D, we’ve managed to reduce costs for users by 20%. This cost-effectiveness, combined with the simplicity and accessibility of our platform, makes Ozone a game-changer.
