Back Back
to to
Stories Stories

Sep 27th 2023

How to implement AI right now

hero-image

Right now there’s incredibly loud push of different AI services in the market. It all started with ChatGPT but it didn’t end with it, Microsoft is promoting their Copilot offering and their competitors have their own revolutionary products they’re selling.

There are some incredible examples of AI already showcased, such as how AI gave paralyzed woman her voice back or how Virgin Voyages created Jen AI, but these are huge investments to start your AI journey.

There is debate on if one should focus on low hanging fruits or major strategic applications of AI, but we would argue that starting small and not forgetting the big picture is still better way to hop on the AI wagon. Investing into a chatbot doesn’t mean you shouldn’t think how AI can revolutionize your business with for example having AI controlling your manufacturing and delivery processes.

Tools for developers

First of all, there’s tools for developers which speed up your daily developer tasks enormously. You can use ChatGPT of course to ask trivial coding hints, but you need to be really careful with its output, it very easily generates unusable code with fake API calls and false methods. What we have found useful is Github Copilot. The code it generates is most of the time accurate and fits to the problem you are trying to solve. You do need to check the code for logical errors though, and possible security problems and you can’t just push it to the production.

Intelligent automation

For your main business you might want to turn your eyes towards things such as intelligent automation, which is area of huge investments right now. Humans are not as quick as AI to do on the fly -optimization of resources based on monitoring, or noticing anomaly patterns from terabytes of machine sensor data.

One could use AI to optimize the supply of raw materials, work and processes based on multiple data sources, which could be very difficult for a person to handle. It is not unheard of to find hundreds of tasks in a company where AI automation would speed up the processes and supercharge your business processes.

Sentiment Analysis

Sentiment analysis is one of the easy areas of AI to take in use at the moment, one could use it to  identify and extract the emotional tone and attitude for example from customer feedback and reviews, and use it to improve their products, services, and customer satisfaction.

Other simple to implement focus areas are document classifications, machine translations and object detection which all can be handled with Microsoft’s Cognitive Services which are very easy to take in use.

Don’t build your house on sand

AI is always based on data. If you throw a massive amount of big data on AI versus lesser amount of more precise data, you might not get better results. For this reason, one of the foundational investments you need to focus on your journey to AI is to have a mature data strategy and platforms.  Investing on data does not only help you with your AI investments but having common standards for data gathering, preparation and organization will most likely open up other opportunities for your business as well.

Came this far? Why not say Hi!

Teemu Tapanila

Teemu Tapanila

CTO, Principal Architect

teemu.tapanila@mallow.fi

+358 452 135 655

Riku Pilli

Riku Pilli

Sales Manager

riku.pilli@mallow.fi

+358 40 725 0888