There’s a guy on our block called Mike – over 80, who spent almost half a century in sales and made all his money before computers really became a thing.
(His stories about the business trips he took to Japan in the 1970s are just insane.)
The reason we’re mentioning Mike is that he is our litmus test for whether something in tech is becoming a thing.
We know this when he asks one of our people, “So, when will they put “this new thing” in my pants?”
For example, “Are they gonna put my pants in a cloud?” or “When are they putting that blockchain in my pants?”
You get it.
Over the last six months, he’s been all about AI and its application to his pants.
Unfortunately, it often feels that people in tech or tech-related fields understand AI as much as Mike does, making them vulnerable to various AI “consultants” and “experts” who peddle snake oil.
Hence, this article.
We want to share, straight-up, what an AI consultant is, what they do, and how they can help your business leverage artificial intelligence for growth and innovation.
What Is AI Consulting?
AI consulting is a service provided by an expert in the AI field who acts as a bridge between artificial intelligence (AI) technology and business strategy. AI consultants guide businesses in identifying, evaluating, and scrutinizing AI opportunities, designing and implementing AI solutions, and ensuring that investments made in this area align with the larger business objectives.
Due to the varied nature of possible AI applications, these larger business objectives can differ greatly, and one of the signs that you are actually working with an AI consultancy that knows what it does is that these objectives will be very clearly defined.
They can include, but are in no way limited to:
- Increasing operational efficiency
- Reducing costs and improving profitability
- Driving innovation and competitive advantage
- Enhancing customer experience and satisfaction
- Optimizing supply chain and logistics
- Improving decision-making processes
- Automating routine tasks and processes
- Enhancing product and service quality
- Identifying new revenue streams and business models
- Strengthening data security and privacy
- Personalizing marketing and sales strategies
- Enhancing risk management and fraud detection
- Improving workforce productivity and engagement
- Streamlining compliance and regulatory processes
- Enabling predictive maintenance and asset management
As you can see, some of these are systemic objectives aimed at benefiting an organization as a whole, while others are focused on specific departments, such as sales, marketing, product, procurement, and so on.
It should be pointed out that while the list of objectives above illustrates the many possibilities of applying AI to your business, they are nothing but an overview and a good starting point for explaining, in more detail, what AI consultants’ responsibilities include.
By the way, if any of the above sounds like something your company might benefit from, give us a call, and let’s figure out if you should build with us.
Responsibilities of an AI Consultant
Arguably, the best way to illustrate the many responsibilities of an AI consultant is to imagine a hypothetical scenario where you are a business owner or a part of its leadership, and you feel like your company has the potential to be improved by implementing an AI solution of some kind.
Perhaps you already have an idea of which aspect of your company this is, or perhaps you just want to explore your options.
You approach an AI consultancy.
The following is what they should do.
Educate you about AI
The biggest red flag you can come across when starting a conversation with an AI consultant is if they pitch you an idea in the first 20 seconds of said conversation. It might sound something like this, “Yeah, sure… We have just the thing you need… It’s an AI-powered thingamajig that makes your business great!”
If you run into someone who opens up with a variant of this, run and don’t look back.
- There is no AI silver bullet that works for every company and instantly makes it amazing.
- There is no marketing AI silver bullet that works for every company and instantly makes it amazing.
- There is no marketing personalization AI silver bullet for Danish e-commerce companies with an MRR of $230,500 and a spectacularly precise market segment.
There may be tools and approaches that might work for you if you run such a company, but there are no guarantees.
Making sure that you understand this is the mark of an expert, conscientious AI consultancy. Depending on how much you understand AI, they need to educate you about its capabilities, limitations, and innumerable intricacies that might influence your company’s adoption of AI.
Ask questions
Once they educate you about AI to a reasonable degree, they should start asking questions.
In this context, asking questions stands for the following:
- Collaborating with you on identifying specific, measurable, and achievable benefits that implementation of AI could have for your business.
- Evaluating your company’s current state, including the functioning of various departments, their interactions, as well as your internal processes.
- Analyzing your industry, competitors, and the general market by drawing from your expertise and experience.
- Conducting feasibility studies to figure out if an investment in AI can produce a high enough return on investment to justify the money that will be spent on it.
- Working with you on building an AI strategy that will clearly define goals, metrics, KPIs, resources, timelines, processes, stakeholders, and risks of what you will be building together or purchasing as a tool.
If, at any moment, you feel they are not asking you enough questions, ask them why this is the case.
This is a collaboration and not a transaction.
Build or implement an AI tool iteratively
Your (carefully) chosen AI consultant/consultancy will either have a team capable of building/implementing the AI solution you agreed upon, or they will have a network of trusted partners that will become involved in this phase.
(If the latter is the case, they should have asked you about whether you are comfortable working with third parties and how you would like to vet them.)
The AI specialists will then go on to do what they do:
- Data collection
- Data preprocessing
- Exploratory data analysis
- Feature engineering
- Model selection
- Model training
- Model validation
- Hyperparameter tuning
- Model evaluation
- Model optimization
- Model deployment
- Monitoring and maintenance
Once again, this is a broad-strokes picture and doesn’t correspond to every AI development process. We just wanted you to understand the scope of the work that is usually included in the tech part of it.
You have probably noticed the word iteratively in the subheading of this section. It denotes an approach to developing and implementing an AI tool in iterations – start with the smallest piece of work that contributes to the end product > analyze it > build on it/improve on it > analyze it > build on it/improve on it > etc.
This is what people often refer to as Agile software development, and it is the way to go, or at least that is the way we do stuff, including building AI functionalities and/or products.
Check out our previous work and see the results of this approach.
When done right, the iterative approach shows a number of benefits when compared to the now-ancient waterfall method of developing products in clearly defined phases.
The most important benefit for us at this point in our hypothetical journey is the ability to regularly inspect if the product is moving in the right direction. Moreover, the fact that you are building it in iterations makes it easier to make course corrections if you and/or your AI consultant start noticing that the product is starting to depart from the original idea as agreed upon when you co-created your AI strategy.
Educate your people
At this point, your AI product/implementation is starting to take shape, and this is the perfect time to start educating the rest of your company about how they can expect to benefit from AI.
Perhaps it will make their jobs easier. Perhaps it will allow them to do things they were previously unable to do. Perhaps it will save them time by automating time-consuming, repetitive tasks. Perhaps it will result in your company’s offering becoming more competitive.
The important thing is that everyone is on board, and your AI consultant is your best ally in this process.
They will know exactly how to communicate AI concepts to non-technical stakeholders and dispel any fears that people might have about being replaced by the evil AI (you would not believe how scared some people are because they’re reading five articles every day telling them how they will be made obsolete by AI in the next four hours).
Moreover, your AI consultants will also educate and train your people on the importance of data governance and ethical AI use. As more and more regulation is being (rightfully) introduced to the AI field, this will become only more and more important.
Manage the rollout
Once again, let’s imagine that up to this point, it’s been nothing but roses. Your new AI tool/functionality/enhancement is completed, and it is all you hoped for. Everyone at your company is up-to-speed and excited to give it a go.
It is time for the rollout.
If you and your AI consultancy have been transparent, careful, and professional up to the rollout, the chances are that things will go well. Minor hiccups may happen (they almost invariably do), but that is why you have an AI consultancy in your corner.
They probably have dozens, if not hundreds, of rollouts behind them, and they should know how to make this process as painless and as non-disruptive as possible. They will probably do it in phases, with a lot of backing up and contingency plans.
Trust them.
But keep talking to them. Keep asking questions.
Perhaps you can help. Perhaps someone on your team can help them.
Almost every problem can be solved with proper communication. Remember that.
And immediately start monitoring and analyzing how your new AI solution is “behaving.”
Are you starting to see the benefits that were clearly defined in the strategy? Does the actual timeline match the one from the strategy? Are your people using the new solution as you expected and as they were trained?
If you are noticing major discrepancies between your expectations and reality, talk to your AI consultancy. Ask them questions. Collaborate on mitigating the issues and optimizing the use of the new tool/feature/solution.
Post-rollout maintenance
As your company is starting to see the expected benefits from your brand new AI solution, your AI consultancy should continue to assist you in maintaining and optimizing the solution, at least for the duration you agreed upon when collaborating on the strategy.
There is always a chance that issues might appear after some use or that you might identify opportunities for further improvements. Your AI consultant should be the first person to get on board once more.
Types of AI Consultants
In the previous (massive) section of the article, we tried to illustrate the responsibilities of a generalist AI consultant or an AI consultancy that provides a holistic service on a hypothetical project that is relatively simple.
The reality, as always, is much more complex, and specialist AI consultants may be called to focus on a particular area. Depending on where you are in your AI journey, you may find yourself in need of their specialized services, so we wanted to cover some of the most common types of AI consultants:
Once again, this is just a selection and not a full list of specialist AI consultants.
Conclusion
Hopefully, this article has provided you with some idea of what AI consultants do and how they can (potentially) help your company. We considered writing a section on how to choose an AI consultancy, but it’s always lame when service providers write stuff like that and obviously recommend themselves.
So we decided to skip that part. (Of course, we’d still like to invite you to take a look at our other services, other than AI consulting, that is)
Just make sure it is someone who gives you straight answers and also asks questions.
For what it’s worth, we ran the article past our friend Mike, and he asked if he now needs an AI consultant for his pants.