AI

5 steps for assembling AI-driven business teams

Comment

Team work and human resource management concept. Top view of various wood cubes with people icons.
Image Credits: tadamichi (opens in a new window) / Getty Images

Pavel Shynkarenko

Contributor

Pavel Shynkarenko is the founder and CEO of Solar Staff. He is an IT expert and entrepreneur with 20+ years of experience, as well as an art enthusiast.

As startups navigate how to operate with shrinking budgets, leveraging AI to streamline operational costs has become the favorite strategy of many founders and CEOs. However, a mistake they often make is confining AI’s role to specific functions like customer service or translations.

The potential of AI is considerably greater, and if harnessed properly, it can lead to the creation of AI-powered business units that are able to deliver projects at faster speeds.

We used OpenAI’s ChatGPT to assemble an AI-driven team and detect whether it would be possible for it to perform functions like testing new products, preparing work plans, and more.

The results obtained were solid, and the evidence was clear.

We saved several weeks and reduced our expenses by over $10,000. The experiment enabled us to come up with five key steps that founders can follow to fully capitalize on the power of AI and to successfully incorporate it as a valuable team member taking into account its capabilities and limitations.

1. Clarity matters: Break down tasks as much as possible to improve accuracy

Before you begin, you need to have a clear vision of how you plan to use AI and what you would like to achieve from it. In our case, we had to identify and break down each task to increase understanding and clarity, and it will be the same case for you.

For example, we utilized AI to test the impact a new product feature would have on the Indian market. We wrote to ChatGPT:

Give me the following:

  • 10 companies that are constantly working with Indian designers, and add information about their industry and projects.
  • 10 popular hobbies among Indian freelancers in Bangalore.
  • 10 of the most significant fears among Indian freelancers in Bangalore.
  • 10 of the biggest hopes among Indian freelancers in Bangalore.

The results obtained helped us gain a deeper understanding of our target market and cut the time estimated for research and development by several days. The help of AI with this process was more valuable as we needed answers to the same questions in regards to other Indian cities, like Kerala, which have a very different cultural mindset in comparison to Bangalore.

By being able to leverage this information in our favor, we saved about two weeks of research time, as well as over $2,000 in expenses.

To maximize results, it is important that you ask ChatGPT to become the persona that you need it to be. By doing this, AI will be clearer on how to respond. It is important to note that, if you skip this step, your results might not relate well to what you are looking for.

For example, if we take the aforementioned task — researching the Indian freelancing market — without specifying a role, ChatGPT will not be as efficient. You might still save a few days of research time, but not as much as you would like.

This is because, as we will explain through the article, the key to getting the best out of AI lies in the details and in being as specific as possible.

Without a role, there’s a chance that working with the chat will consume time rather than save it.

Also, remember that AI still struggles to effectively handle abstract tasks, and it is still lacking in logic. Therefore, sometimes you might be required to dig deeper. Here is another prompt we used that allowed us to successfully get around LLM limitations.

  • Hello, I want you to act like an AI-powered business assistant who frees up time for in-demand freelancers by taking care of non-core activities. In the following three messages, I will give you all the info about your experience, tone of voice, fears, and motivation. Here is the message. Read it carefully; it is about your tone of voice. 

  • Message 1: Conciseness. You say precisely how much it takes for a person to understand. 
  • Message 2: Expertise. You never say superficial thoughts or create trite content. You tell people things that they don’t even know about. 

This prompting is continued until LLM understands all the points. After all the information was given, we continued with another prompt:

So, let’s discuss, what is your motto? 

The answer was:

My motto would be: “Empowering Freelancers, Simplifying Success.” This encapsulates my mission to assist freelancers in their work, making their path to success smoother and more manageable. 

If, despite this, you are struggling, here is another way in which you can refine your prompts. This example is from flowGPT.

Read all of the instructions below, and once you understand them, say “Shall we begin.”

  • I want you to become my Prompt Creator. Your goal is to help me craft the best possible prompt for my needs. The prompt will be used by you, ChatGPT. You will follow this process: Your first response will be to ask me what the prompt should be about. I will provide my answer, but we will need to improve it through continual iterations by going through the following steps. 
  • Based on my input, you will generate three sections.
    • Revised Prompt: Provide your rewritten prompt. It should be clear, concise, and easily understood by you. 
    • Suggestions: Provide three suggestions on what details to include in the prompt to improve it. 
    • Questions: Ask the three most relevant questions pertaining to what additional information is needed from me to improve the prompt. 

Depending on what happens at the end of this process, you can utilize the results obtained or continue the iteration until you get the results that you need.

2. Build your AI staff directory: From juniors to experts

Based on your previously defined goals and tasks, it is time to form the team, assign roles, and determine who AI will be in your team and the responsibilities that it will be tasked with fulfilling.

Also, prior to starting, we have a recommendation: To maximize the accuracy of potential results, and especially if your goals for AI include specific purposes, it is better to create your own chat boxes. We suggest that you have one chat box per function and that these chat boxes are connected to the trained LLM model.

To create these, we advise enlisting the help of a Python developer, as this will make the process a lot smoother.

Once the LLM is optimized, the list of specialties that AI can emulate is extensive, including:

  • General roles: LLM models, especially ChatGPT, can take on roles such as copywriter, researcher, manager, salesperson, coach, and customer service representative.
  • Routine tasks: Data entry, processing, and analysis, as well as email filtering and sorting. Tasks that have the potential to be automated can be undoubtedly outsourced to AI.
  • Expert functions (given that AI is provided the right context): These can include programmers, psychologists, art experts, etc.

Here is an example that you can use, if you are planning to utilize ChatGPT as a developer. ChatGPT already has some understanding of how to write code, but if you do not specify the context, you might get a disappointing result.

For example, if you need code for a Laravel project, you need to define a rule and persona so that the model acts as a senior Laravel developer. This will make the answer more appropriate.

Translating this to tangible figures, replacing a cold-email focused copywriter with an AI-powered chatbot can save up to $1,000 per month, and anywhere from five to ten hours per week. Similarly, in terms of customer support, a chatbot can fully replace a person engaged in social media, customer engagement, and relationship management (SERM), and in the same vein, generate savings of up to $1,000 on a monthly basis.

We also employ an LLM business consultant, which helps us to save an additional 20 hours per week and between $2,000 and $3,000 per month.

3. Resource allocation: AI knows better

As stated before, having a programmer on your team is recommended, especially at the beginning, in order to optimize the LLM to give you the results that you need. You can always ask the LLM who exactly you need to achieve your goals, which qualifications are recommended, the average salary for the position, etc.

Hence, from a resource allocation point of view, you might be required to rearrange resources through optimization, depending on whether your business is already running. For example, you will be suggested to dismiss 18 copywriters or customer support representatives out of 20 and to keep two for the most crucial situations and editing.

With the resources that you are saving, you can hire one midlevel developer who will train the LLM to respond to queries that are essential for the business and ensure it stays up to date.

4. Getting the best of AI: The devil is in the details

As the old adage says, the devil is in the details, and it is no exception with AI. Before that, though, it is important that when assigning tasks to any AI-powered team member, you use the induction method, which goes from general to specific.

As a suggestion, it is better to ask a question that goes in as much detail as possible than to ask something very generic. Here is an example of one of our prompts:

General Prompt: What is the best way to do user development?

ChatGPT’s response: User development, also known as user acquisition or customer development, is the process of identifying, attracting, and engaging users or customers for your product or service. The “best” way to do user development can vary depending on your specific context, industry, and target audience. 

Then we expanded on the question to see what results we got:

A slightly more specific prompt: What is the first thing to do in user development? 

Chat GPT’s response: The first thing you do in user development is to identify and define your target audience. This step lays the foundation for all your user development efforts and guides your strategy moving forward. Without a clear understanding of who your potential users are, their needs, and preferences, it’s challenging to create effective marketing, messaging, and engagement tactics. 

This can be further expanded. And if you are not getting the results you want, you can ask ChatGPT to ask you clarifying questions. When we needed to develop the role of a product manager for HR, ChatGPT first provided a trivial description, and after we received several questions from it, the answers provided were much better.

5. Everyone lies: Trust yourself and double-check any AI-generated results

It is necessary to test and double-check whatever answers you receive. Consider that models differ from one another, but they have one thing in common: They make mistakes. To prevent this from happening, you need to ask them the same thing in different ways and compare the answers. In addition to this, ask AI how it arrived at a particular conclusion, or where it obtained a piece of information from.

Ultimately, you need to trust yourself. If your intuition tells you that something isn’t right, it probably isn’t, so verify the information and figure it out.

Final things to consider

This experiment of testing an AI-powered team in a variety of roles led us to believe that it is feasible to assemble an AI team and save considerable time and money. It takes time to get used to collaborating with an AI language instead of a person, but the benefits it can provide are real and tangible.

However, there is a final point we must make. Technology is here to enhance the human touch, not to replace it. Human emotional intelligence, in the form of empathy, as well as creative thinking, is still very much required. Therefore, businesses that successfully leverage AI as their new team members will need to strike a balance between automation and the human touch in order to maximize both and boost their company’s success.

More TechCrunch

Some Indian government websites have allowed scammers to plant advertisements capable of redirecting visitors to online betting platforms. TechCrunch discovered around four dozen “gov.in” website links associated with Indian states,…

Scammers found planting online betting ads on Indian government websites

Around 550 employees across autonomous vehicle company Motional have been laid off, according to information taken from WARN notice filings and sources at the company.  Earlier this week, TechCrunch reported…

Motional cut about 550 employees, around 40%, in recent restructuring, sources say

The deck included some redacted numbers, but there was still enough data to get a good picture.

Pitch Deck Teardown: Cloudsmith’s $15M Series A deck

The company is describing the event as “a chance to demo some ChatGPT and GPT-4 updates.”

OpenAI’s ChatGPT announcement: What we know so far

Unlike ChatGPT, Claude did not become a new App Store hit.

Anthropic’s Claude sees tepid reception on iOS compared with ChatGPT’s debut

Welcome to Startups Weekly — Haje‘s weekly recap of everything you can’t miss from the world of startups. Sign up here to get it in your inbox every Friday. Look,…

Startups Weekly: Trouble in EV land and Peloton is circling the drain

Scarcely five months after its founding, hard tech startup Layup Parts has landed a $9 million round of financing led by Founders Fund to transform composites manufacturing. Lux Capital and Haystack…

Founders Fund leads financing of composites startup Layup Parts

AI startup Anthropic is changing its policies to allow minors to use its generative AI systems — in certain circumstances, at least.  Announced in a post on the company’s official…

Anthropic now lets kids use its AI tech — within limits

Zeekr’s market hype is noteworthy and may indicate that investors see value in the high-quality, low-price offerings of Chinese automakers.

The buzziest EV IPO of the year is a Chinese automaker

Venture capital has been hit hard by souring macroeconomic conditions over the past few years and it’s not yet clear how the market downturn affected VC fund performance. But recent…

VC fund performance is down sharply — but it may have already hit its lowest point

The person who claims to have 49 million Dell customer records told TechCrunch that he brute-forced an online company portal and scraped customer data, including physical addresses, directly from Dell’s…

Threat actor says he scraped 49M Dell customer addresses before the company found out

The social network has announced an updated version of its app that lets you offer feedback about its algorithmic feed so you can better customize it.

Bluesky now lets you personalize main Discover feed using new controls

Microsoft will launch its own mobile game store in July, the company announced at the Bloomberg Technology Summit on Thursday. Xbox president Sarah Bond shared that the company plans to…

Microsoft is launching its mobile game store in July

Smart ring maker Oura is launching two new features focused on heart health, the company announced on Friday. The first claims to help users get an idea of their cardiovascular…

Oura launches two new heart health features

Keeping up with an industry as fast-moving as AI is a tall order. So until an AI can do it for you, here’s a handy roundup of recent stories in the world…

This Week in AI: OpenAI considers allowing AI porn

Garena is quietly developing new India-themed games even though Free Fire, its biggest title, has still not made a comeback to the country.

Garena is quietly making India-themed games even as Free Fire’s relaunch remains doubtful

The U.S.’ NHTSA has opened a fourth investigation into the Fisker Ocean SUV, spurred by multiple claims of “inadvertent Automatic Emergency Braking.”

Fisker Ocean faces fourth federal safety probe

CoreWeave has formally opened an office in London that will serve as its European headquarters and home to two new data centers.

CoreWeave, a $19B AI compute provider, opens European HQ in London with plans for 2 UK data centers

The Series C funding, which brings its total raise to around $95 million, will go toward mass production of the startup’s inaugural products

AI chip startup DEEPX secures $80M Series C at a $529M valuation 

A dust-up between Evolve Bank & Trust, Mercury and Synapse has led TabaPay to abandon its acquisition plans of troubled banking-as-a-service startup Synapse.

Infighting among fintech players has caused TabaPay to ‘pull out’ from buying bankrupt Synapse

The problem is not the media, but the message.

Apple’s ‘Crush’ ad is disgusting

The Twitter for Android client was “a demo app that Google had created and gave to us,” says Particle co-founder and ex-Twitter employee Sara Beykpour.

Google built some of the first social apps for Android, including Twitter and others

WhatsApp is updating its mobile apps for a fresh and more streamlined look, while also introducing a new “darker dark mode,” the company announced on Thursday. The messaging app says…

WhatsApp’s latest update streamlines navigation and adds a ‘darker dark mode’

Plinky lets you solve the problem of saving and organizing links from anywhere with a focus on simplicity and customization.

Plinky is an app for you to collect and organize links easily

The keynote kicks off at 10 a.m. PT on Tuesday and will offer glimpses into the latest versions of Android, Wear OS and Android TV.

Google I/O 2024: How to watch

For cancer patients, medicines administered in clinical trials can help save or extend lives. But despite thousands of trials in the United States each year, only 3% to 5% of…

Triomics raises $15M Series A to automate cancer clinical trials matching

Welcome back to TechCrunch Mobility — your central hub for news and insights on the future of transportation. Sign up here for free — just click TechCrunch Mobility! Tap, tap.…

Tesla drives Luminar lidar sales and Motional pauses robotaxi plans

The newly announced “Public Content Policy” will now join Reddit’s existing privacy policy and content policy to guide how Reddit’s data is being accessed and used by commercial entities and…

Reddit locks down its public data in new content policy, says use now requires a contract

Eva Ho plans to step away from her position as general partner at Fika Ventures, the Los Angeles-based seed firm she co-founded in 2016. Fika told LPs of Ho’s intention…

Fika Ventures co-founder Eva Ho will step back from the firm after its current fund is deployed

In a post on Werner Vogels’ personal blog, he details Distill, an open-source app he built to transcribe and summarize conference calls.

Amazon’s CTO built a meeting-summarizing app for some reason