Want to create a AI Product or Service for your Business? Key things to start on your Business AI Journey..

What are your goals? If you are looking for Image classification, Recommending services or product, Customer lifetime modeling to retain your most profitable customers, Churn modeling, Dynamic Pricing to set prices based on demand, Spam detection, other use cases based on detecting patterns in business data points — you will need what is called supervised Machine Learning.

Selecting a AI Annotation Strategy and Tool will be critical for success.

Lets start with some basics and proceed from there.

defines Artificial Intelligence, AI, as “the ability of a machine to perform cognitive functions we associate with human minds, such as perceiving, reasoning, learning, interacting with the environment, problem solving, and even exercising creativity. Examples of technologies that enable AI to solve business problems are robotics and autonomous vehicles, computer vision, language, virtual agents, and machine learning.”

Foundational Facts of AI

Artificial Intelligence (AI), is essentially software used to simulate human intelligence processes, including: learning, analyzing data and reasoning.

As a subfield of AI Machine Learning, ML, uses numerical techniques from computing, optimization, and statistics to “learn” to perform tasks using algorithms instead of being programmed directly from start to finish. Relying on patterns and inference, ML algorithms build a mathematical model based on sample data, known as “training data”.

Uses of Machine Learning Analytics

Descriptive: Understand patterns in data, such as for KPIs and company performance;

Predictive: Used to analyze risks and opportunities for companies based on historical data and statistical models and ML;

Prescriptive: Gartner defines as a “form of advanced analytics which examines data or content to answer the question “What should be done?” or “What can we do to make _______ happen?”, and is characterized by techniques such as graph analysis, simulation, complex event processing, neural networks, recommendation engines, heuristics, and machine learning.”

Types of Machine Learning

Supervised Learning: An algorithm uses training data and feedback from humans to learn the relationship of given inputs to a given output.

Unsupervised Learning: An algorithm explores input data without being given an explicit output variable (eg, explores customer demographic data to identify patterns)

Reinforcement Learning: An algorithm learns to perform a task simply by trying to maximize rewards it receives for its actions (eg, maximizes points it receives for increasing returns of an investment portfolio)

I am going to focus on Supervised learning, where humans are a part of the process in labelling data and defines the output variable.
ML algorithm is trained on the data to find the connection between the input variables and the output.
Once training is complete–typically when the algorithm is sufficiently accurate–the algorithm is applied to new data

How do you do label data?

Data annotation is the process of labeling data to show the outcomes desired from the ML model of datasets with features the ML system should recognize on its own — through: labeling, tagging, transcribing.
There are various annotation tools available. Start with the Business objectives and resources to create the AI Strategy to minimize loss of resources and gain the value of AI.

Important points in the selection of tools:
1. Start with a Proof of Concept — OutSecure team can help create an AI Framework to Build AI Products by creating a proof of concept in a week. This will help determine if your business problem is a good fit for Supervised learning Machine Learning.

2. Data types — Videos or images, text;

3. Build or Buy — Who is going to use the tool? Skilled resources and capability is key for this decision. Buying a tool is a cheaper option but will lack customizing capability and that may be an important tradeoff for the business outcomes.

4. Opensource vs Commercial tools — These are the same considerations when it comes to security and support as in traditional business use. There are a lot of communities with opensource tools, one we are members of is The Linux Foundation. If you have development resources opensource tools are a good option. Access to skilled resources is key if you choose the open source option.

5. Security — Security is a multi dimensional area in traditional systems but when it comes to AI there is a complex dynamic that goes across security of data, Privacy of data and algorithmic integrity to prevent building systems that lack Trust in the outcomes. Data Annotation tool and datasets must be carefully examined for regulations such as GLBA, HIPAA, GDPR at a minimum.

You can read more about my AI Trust — Call to action.

Pamela Gupta, President of OutSecure Inc, a woman owned Trusted AI strategy creation company. Pamela has been Advancing Trust in AI through Security, Privacy, Transparency and writes on AI Strategy for Organizations.

“ I learned a lot on AI Product Management from

in her Artificial Intelligence Bootcamp for Product Managers & Business Managers” at Stanford Continuing Studies and from her AI Ethics course at Business School of AI.”