• karlduane

Business Applications of Existing AI technology

Last week I had a conversation with the CTO of a small company (10-50 people) about how the company uses its data and a portion of the conversation has stuck with me ever since. In the conversation of machine learning, AI, and Data Science we often hear about the large scale breakthroughs, the newest iteration and gigantic undertakings of titans. GPT-3 anyone? While anyone could theoretically be the next big innovator and discover the core details that could create a tool that can rival GPT-3, the large teams and scope of the undertaking aren't right for every company.

But what about the startup that wants to use existing tools? For some companies, it is more cost effective to integrate existing open source tools or API's into their tech stack.

A few ideas that I have run across just for toolsets I am familiar with:

  • A computer vision model could be paired with a simple webcam to identify pests or blight in a garden or on a farm and can be used to recommend a course of action

  • Optical Character Recognition tools are widely available, implement a pre-trained model with bounding boxes to create a smart scanning tool for handwritten forms. Turbotax demonstrates this with their quick tax form scanning tool, and it could also be implemented for a state or municipal tax forms that are not currently supported by big companies

  • Law firms can use Hugging Face NLP transformers to scan and tokenize legal documents such as legislation and pair it with a knowledge graph tool such as Grakn to allow the firm to evaluate how proposed legislation will affect their clients.

  • The same pairing of language transformation and knowledge graphs could be applied to proposed thousand-page legislation that passes through a legislator's office and more quickly identify how it affects issues they are focused on.

As a data scientist, I add value by helping companies understand how machine learning and AI technologies fit into their objectives and key results or key performance indicators. Where does your data trail lead? Is it where you want to go?

3 views0 comments

Recent Posts

See All

Recently I had the opportunity to attend the LinkedIn DATAcated conference. It included a number of excerpts from big names in the data science world and I wanted to record some of my favorite quotes,

#celebrated_failures I've been spending the last two weeks reworking a group project into something I can deploy. The group decided to build a Covid-19 Misinformation Classifier that could give a qui