WIDE IO: algoBox
The eBay for algorithms, algoBox is a cloud based algorithm exchange marketplace that helps businesses find and use the best and the latest algorithms. These algorithms are contributed by researchers, scientists, technology enthusiasts and business labs, and as such algoBox serves as a bridge between the research community and businesses.
Algorithms are the brains of the software. Algorithms refine complex data into simple conclusions and smart decisions. From stock markets to traffic lights, algorithms are used everywhere, and in today's digital age, using the right algorithm is the key driver of success for digital goods and services.
algoBox enables consumers (businesses) to find and use the best algorithms quickly, efficiently and in a cost-effective manner. Algorithms are provided as APIs that are easy to integrate into any code and follow a pay-as-you-go model.
For producers, i.e. researchers, scientists, technology enthusiasts and business labs, algoBox provides a platform to commercialise their work and earn revenues from their research.
A number of good algorithms are trapped in labs and research papers. These algorithms have huge potential and a wide range of applications. Designed by experts who spend a large amount of time perfecting them, these algorithms can solve various problems ranging from effective traffic management to predicting crime using suspicious behaviour, and in doing so earn revenues for the researcher. With algoBox, we aim to provide efficient usage and distribution of the algorithms.
Building algorithms is a time consuming process and requires a lot of resources and technical expertise. Engineers at Amazon and Google spend huge amounts of time trying to build new algorithms. But what about companies who do not have the resources similar to that of Microsoft or Facebook? They usually build algorithms that are not optimal or outsource the job. But outsourcing is inefficient, slow, costly and the algorithms are very quickly outdated. The scope of the algorithm and level of expertise is also limited to the outsourced company and the output is not always satisfactory.
1) Better, Efficient Software
algoBox allows businesses to create more efficient software by ensuring that they have access to the best algorithms. We provide businesses with a pool of algorithms contributed by a community of experts.
2) Reduced Costs
algoBox follows a pay-as-you-go model. So businesses only pay for what they use. Unlike outsourcing, businesses do not have to commit to any one algorithm but can quickly move to the most advanced algorithm available for their job.
3) Faster Development
Algorithms are provided as APIs that can be easily integrated into any code. Our standardisation technology could potentially reduce software development time by as much as 2 weeks.
Researchers require funds to support their research. algoBox provides the research community with not only a way to commercialise their research, but also a way to test the application of their technology and generate revenues.
1) Commercialisation of research
The greatest reward for any researcher is to see their technology in use. algoBox provides the research community with a way to ensure that research finds application.
2) β-Tests, Hosting
We take care of hosting the algorithms and computational needs so researchers don't have to worry about the server. They can also test the application of their technology and improve based on the response and review from the users on the platform.
We give researchers a platform to earn revenues from their research.
Substantial accomplishments to date
We've been actively working on algoBox actively and our company has received a number of awards and recognition. Some of the company's accomplishments are:
1) AUG 2014: The company received a £20K grant from TSB.
2) JUL 2014: 199 Algorithms on the beta version of algoBox.
3) DEC 2013: Part of Rackspace Startup: £15K savings on server space.
4) DEC 2013: London Fusion Traction Award of £10K to work on algoBox.
5) OCT 2013: We've been a part of Growth Accelerator Program and have successfully tested the product with a number of developers.
6) JUN 2013: Creation of Minimum Viable Product/Prototype: Beta version released.
7) FEB 2013: Part of Microsoft BizSpark, free access to most Microsoft applications.
algoBox follows a brokerage model. It'll serve as a platform that connects businesses with algorithms and charge a fixed percentage (commission) on each transaction and a base fare.
Algorithms would be provided in the form of APIs. API or Application Programming Interface (API), specifies how software components should interact with each other. APIs are easy to integrate into any code and businesses would be charged based on the number of times an API has been called (i. e. the algorithm has been used). For example, if the price of calling an API of a certain algorithm is £100, we take £40 as our commission and £60 goes to the producer.
On an average, we expect to make £100 for a good algorithm each month.
The price of calling an API is set by the producer but we assist them in setting up the right price to maximise the consumer and producer surplus. However, this price may vary based on the time, efficiency, complexity and usage of the algorithm. Our revenue increases with the number and quality of algorithms available on the platform and so we'll be trying to not only increase the number of algorithms available on algoBox, but also the quality and application of those algorithms.
Given that we'll have a significant amount of data, the company may be able to prepare reports on trends and earn revenues from these reports.
We're not just a library of APIs where the producer is responsible for everything. We're easy to use and efficient as we take care of the server needs for running the algorithms. This reduces costs for the producers, makes it easy for them to be on the platform and should give us a reputation for providing a standard and reliable service.
Use of proceeds
The investment will be used to build on the prototype and to create the first public version of our platform algoBox. We've been improving the prototype and have been in constant touch with producers and consumers. We've identified the unmet need and have a clear roadmap to build a successful product.
1) Hire senior software engineer (x1) and senior web developer (x1) to build the first public version of algoBox.
2) Market the product and add more producers to the platform.
3) Prepare the organisation for a VC round.
September / October 2014: Secure Investment.
September 2014: Hire Senior Software Engineer (x1) and Senior Web Developer (x1).
November 2014: Marketing campaigns to add more producers and generate awareness.
December 2014: Pre-launch; debugging and improvements.
February 2015: First public version
algoBox operates in 2 markets:
1) Academia (this is where most of our producers come from)
2) Analytics (Video and NLP)
algoBox would begin operations by focusing on the niche segment of the analytics market i.e. Natural Language Processing (NLP) and Video (including Image) Analytics. NLP is concerned with the interactions between computers and human (natural) languages. For example: the computer understands that "I ate an apple" doesn't relate to 'iPhone'. Video Analytics, is a significantly growing segment of the market and involves predictive analytics, video security, augmented reality and multimedia.
We already have 199 algorithms on the platform including algorithms to MEASURE TIREDNESS, RECOGNISE EMOTION, DETECT UNATTENDED OBJECT etc.
algoBox has applications in:
1) Security: What if YouTube and Facebook could detect illicit content as soon as it’s uploaded? And how about facial recognition on the move? Organisations can stay up to date with the latest technology on video and image security, privacy and safety.
2) Multimedia: Imagine if you could create a billboard (advertisement) that interacts with consumers and plays ads based on the customer's mood. This involves live behaviour (video) analysis, ad processing and feedback. algoBox would enable multimedia firms to produce next-gen ads that are more creative.
3) Analytics: How about consumer behaviour analysis, live from the supermarket? Analytics and consultancy firms could use algoBox for the latest and up-to-date algorithms for analysis and recommendations: both NLP and Video. We are also working towards providing a trusted data environment.
Characteristics of target market
We've observed that marketplaces lead to expansion of the market and in doing so capture the most profits; e.g. eBay makes more money than individual sellers/stores. algoBox's business model is similar to that of eBay; connecting buyers with sellers and controlling the interdependent links in the value chain (likely to capture most profits).
Video analytics is seeing an enormous growth and MarketsandMarkets predicts that while the video surveillance applications market should reach $25.3 Billion by 2016, Intelligent Video Analytics (security and business intelligence) market in 2017 will be worth $867.8 million, up from $180.0 million in 2012. Natural Language Processing is expected to be worth $9.8 Billion by 2018.
With an increasing interest and use of APIs and significant growth of our target market, we feel that this is an ideal time to launch algoBox. The adoption rate is high and 5 Forces Analysis reveals that the industry is attractive.
Before we approach businesses for sales, we need to get producers on board. Our goal will be to build a collection of algorithms first and then reach out to potential businesses.
1) Network: We feel that we have strong relationship with the research community. Our founder has a PhD and is well known in the research community. We've tested the private beta of algoBox with various researchers and have carried out numerous demonstrations at universities in the UK. Universities have shown strong interest in the platform and we plan to build on it for marketing.
2) Events: Most of our contacts with the research community are made at events. Conferences, trade shows, product demonstrations etc. all give us the opportunity to personally contact the researcher and get them on board. The research community is very close and we are well positioned to benefit from it and use word of mouth as a way of promotion ("you're not using Wide IO? The professor next door earned 5K last month from it!").
1) Events: Trade shows, products demonstrations etc. attract businesses looking for the latest technologies. Events are great for meeting potential customers. We not only meet the consumer, but also understand his needs, see trends etc allowing us not only to add consumers, but also improve platform based on consumer preferences.
2) Producers: It is in the interest of the producer to bring in more consumers to the platform as his revenues are directly related to the sale of their algorithms. As such, we think the producers are likely to promote the platform and even give details of their algorithms on the platform in publications. Universities would also be incentivised to promote the platform.
+ Standard Report/Magazine: We are considering an online magazine/report to educate our customers on latest trends; including information like what kind of algorithms are most popular, what are the latest algorithms on the platform, events etc.
It is not about building the best platform, it's about building the right one. algoBox is a platform that empowers users to create value and in order to stay on top of the ladder, we'll use the following factors:
1) HIGH ENTRY BARRIERS The industry has high barriers to entry (developing the platform is difficult), reducing the threat of new entrants. In order for the current competitors to operate in our niche, they would have to give up their current customers, which is not feasible and highly unlikely. We do intend to diversify, but we want our growth to be steady and sustainable.
2) PATENT While competition is inevitable (and somewhat beneficial), we do realise that we are not the first movers in the 'Algorithm as a Service' space. But unlike our competitors, we're focused on a niche and plan to be the biggest player of our target market. We're also working on a standardisation technology that we believe will give us sustainable competitive advantage. We have full faith in our product and intend to patent the technology as soon as possible.
3) NETWORK EFFECTS As more and more producers (research community) join algoBox, we'll have strong network effects and should be able to defend our market position.
4) CUSTOMER RELATIONSHIP We would be able to capitalise on our strong relationship with the research community. We've tested the private beta of algoBox with various researchers and have carried out numerous demonstrations at universities in the UK and universities have shown strong interest in the platform.