Development of AI Products from Concept to Market
On May 6th, 2019, students in the Artificial Intelligence: From Concept to Market MBS course presented their final projects under the guidance from Prof. Richard Mammone. The students had developed technology as well as devised plans to enter the market as a start-up company. Five teams used various AI technologies such as CNN, transfer learning, and LSTM to develop their products:
- Team Celebrand developed an application that discovers the brands of accessories worn by celebrities in a given image. The major breakthrough here is finding the brand name even though no logo appears in the image (or on the item).
- Team Prognosis developed an application that detects melanoma in the skin. When given an image or text related to skin, the application returns a diagnosis as to whether the patient has skin cancer or not.
- Team Sentix developed an application which predicted the sentiment of a company as positive, negative or neutral based on the tweets related to the company on Twitter.
- Team Skin Scan developed an application to detect 7 types of skin problems. When given an image, the application determines if there is any problem with that patient. If a problem is identified, the application checks it against 7 unique types of skin problems.
- Team Theia developed a helper application for visually impaired people. When an object, such as a bottle or a switch is photographed using this app, the app gives an audio message stating the name of the object. This app can use be used in real-time and regular phone cameras to capture images.
Presentation by the Celebrand team
Presentation by the Prognosis team
Presentation by the Skin Scan team
Presentation by the Sentix team
Presentation by the Theia team
The presentations covered a complete analysis of the market and introduced the products to the audience. Each team was considered as a separate start-up company. They prepared competitive analysis, financial projections for 5 years, the timeline for the company, the IP’s which are close to their product, business model canvas, their value proposition, SWOT analysis, Porter’s 5 forces analysis, pricing model, and marketing strategies. They had planned and worked out the potential influencers for their product, potential acquisitions, and sources for their revenue streamline. Various strategies to generate revenue, including search engine optimization, revenue through Google ads, cost per click, and an AdWords campaign, were discussed.
In conclusion, the students were able to combine their AI skills with business acumen to develop products that have real-world benefits. The audience enjoyed both the highly informative presentations and the delicious pizza that accompanied the presentations.