Back to Blog

How Tweet Hunter scaled to an 8-figure exit with AI21’s LLM

,
,
,
February 21, 2023
No items found.

Discover how Tweet hunter built its product and eventually landed an 8-figure exit with AI21 Studio.

The Brand 

Tweet Hunter is an all-in-one Twitter growth tool, designed to help users grow and monetize their Twitter audience. Their goal is to make it as easy as possible for users to create high-performing content, build an audience around their topics of expertise, and monetize opportunities.

notion image

The Story 

The founders of Tweet Hunter started with an insane challenge: ship one new product every week till you find the one that fits. 

“Generating revenue from the product was our number one, clear signal for validation.” says Thibault Louis-Lucas, founder of Tweet Hunter. 

The team started with Twitter as a distribution channel for their products because they had an engaged following of 2,000 people. So they started building a product to help people with a Twitter following generate sales. 

Their principle was: consistently great content was crucial to a creator’s growth on Twitter. 

But they needed an LLM partner who could see their vision and fine tune for their specific use-cases. 

notion image

Why AI21? Shared core values, flexibility, and excellence

Tweet Hunter was looking for ways to empower creators to write incredible content faster, by removing gruntwork from the equation. They wanted to create tools that help creators focus on the creative aspects instead of the mundane tasks like editing, collection and formatting. Tweet Hunter's product needs were perfectly aligned with AI21's mission: to help developers build AI-first writing experiences.

Switching from the legacy GPT-3, Tweet Hunter found this vision only in AI21. 

“We could’ve switched back to GPT-3 but we stayed because of the flexibility that AI21 was providing with the fine tuning of models.” says Thibault.

One of the bigger differentiator for Tweet Hunter was AI21’s willingness to create a customized pricing plan to sync with Tweet Hunter’s usage and growth. Tweet Hunter saved significantly on customization costs because AI21’s pricing for custom fine-tuned model usage is the same as the foundation model usage. Other LLMs (such as Open AI’s) offer customization at approximately six times the cost of their foundational models. 

With these criteria coming together perfectly, Tweet Hunter set out to optimize the product to have an 8-figure exit. 

The first challenge: Generative AI for social media 

Social media channels have strict specifications such as character limits, use of hashtags, and even tone of voice which makes effective generative AI in social media a challenging task, requiring intense experimentation and refinement. 

Initially, this was a challenge for Tweet Hunter since their primary product offering was designed for Twitter. 

AI21 offered a 3-click custom model training which Tweet Hunter effectively leveraged using their proprietary data to fine-tune a model for the exact capabilities they envisioned. AI21 were flexible with both their approach and features so Tweet Hunter could build a tool with multiple junctions for AI to assist.

The second challenge: too many distinct use cases

Writing engaging tweets presents a multi-level challenge: start with a creative hook, keep it succinct, and build an engaging experience for threads. 

This means your AI text generator must be adaptable to different use cases. Also, each of these use cases need to integrate seamlessly while being mindful of the user experience. 

“Twitter’s own user interface and algorithm is optimized for engagement and consumption, it's not optimized to inspire you. And that’s a problem” says Thibault.

AI21 anticipated diverse use cases and built their large language model to be flexible and adaptable from the onset. As a result, Tweet Hunter could fine tune the model in three clicks, iterate quickly and reach success. 

The resulting model was built with three distinct capabilities and two sub-capabilities 

  • Thread idea generator
  • Hook generator
  • Tweet writer

Sub-capabilities

  • Tweet extender (expands on the tweet you’re already started writing)

Each of these features are presented in a fluid transition on the Tweet Hunter platform, which lead to over 5000 paying customers for the tool.

The third challenge: unreliable data to train large language models 

Tweet Hunter’s founders had tried every method under the sun to consolidate relevant data, from using freelancers to creating spreadsheets themselves. 

Not only was this exercise time intensive but it also produced inconsistent and unreliable results. 

They were so disappointed after trying out multiple generative AI tools that they switched back to manual data collection. They were worried that poor quality examples would interfere with the very premise of Tweet Hunter: inspiration for great content. 

Enter AI21. 

The studio’s in-built functionality allowed Tweet Hunter to fine tune the model to learn from a diverse tweet database. AI21's generated suggestions became exponentially better as the system evolved, and Tweet Hunter's goal of helping creators write save-worthy tweets was achieved in a fraction of the time expected.

AI21’s custom model and proactive engagement throughout the journey was exactly what Tweet Hunter needed in a partner. 

“We deeply appreciate AI21’s level of involvement with our product and needs in the early days. We were able to evaluate the model’s quality before production and maintain high standards on the quality of tweets.” says Thibault. 

What changed for Tweet Hunter after working with AI21?

In the last two years, Thibault’s personal account grew from 2,000 followers to 60,000 followers. With Tweet Hunter's scheduling feature and tweet generation tools powered by Ai21’s language models, he (and other users) could tweet frequently and effectively. 

Tweet Hunter scaled to 1M ARR and an 8-figure exit in under a year. 

“The content is very high quality. The tool works with diverse creators, varied niches, and complex topics. So most of the growth comes from the fact that our users are actually successful with the tool.” says Thibault

The level of personalization possible with AI21 propelled Tweet Hunter to first place for multiple creators. 

https://twitter.com/mvxlondon/status/1621192750900051971

“Personalized, highly viral tweet formats developed with AI21 helped busy users share valuable content. That was the win. Our tool became 10 times faster with the AI integration.” says Thibault. 

The road forward

With lempire’s acquisition of Tweet Hunter, the road forward is fast-paced and fascinating. 

The team plans to introduce new features to the platform based on public feedback. AI21 is also set to be integrated with other products in lempire’s toolkit: personalized cold emails, automated follow-ups, and engagement with leads across multiple channels.

Discover more

What is a MRKL system?

In August 2021 we released Jurassic-1, a 178B-parameter autoregressive language model. We’re thankful for the reception it got – over 10,000 developers signed up, and hundreds of commercial applications are in various stages of development. Mega models such as Jurassic-1, GPT-3 and others are indeed amazing, and open up exciting opportunities. But these models are also inherently limited. They can’t access your company database, don’t have access to current information (for example, latest COVID numbers or dollar-euro exchange rate), can’t reason (for example, their arithmetic capabilities don’t come close to that of an HP calculator from the 1970s), and are prohibitively expensive to update.
A MRKL system such as Jurassic-X enjoys all the advantages of mega language models, with none of these disadvantages. Here’s how it works.

Compositive multi-expert problem: the list of “Green energy companies” is routed to Wiki API, “last month” dates are extracted from the calendar and “share prices” from the database. The “largest increase“ is computed by the calculator and finally, the answer is formatted by the language model.

There are of course many details and challenges in making all this work - training the discrete experts, smoothing the interface between them and the neural network, routing among the different modules, and more. To get a deeper sense for MRKL systems, how they fit in the technology landscape, and some of the technical challenges in implementing them, see our MRKL paper. For a deeper technical look at how to handle one of the implementation challenges, namely avoiding model explosion, see our paper on leveraging frozen mega LMs.

A further look at the advantages of Jurassic-X

Even without diving into technical details, it’s easy to get a sense for the advantages of Jurassic-X. Here are some of the capabilities it offers, and how these can be used for practical applications.

Reading and updating your database in free language

Language models are closed boxes which you can use, but not change. However, in many practical cases you would want to use the power of a language model to analyze information you possess - the supplies in your store, your company’s payroll, the grades in your school and more. Jurassic-X can connect to your databases so that you can ‘talk’ to your data to explore what you need-  “Find the cheapest Shampoo that has a rosy smell”, “Which computing stock increased the most in the last week?” and more. Furthermore, our system also enables joining several databases, and has the ability to update your database using free language (see figure below).

Jurassic-X enables you to plug in YOUR company's database (inventories, salary sheets, etc.) and extract information using free language

AI-assisted text generation on current affairs

Language models can generate text, yet can not be used to create text on current affairs, because their vast knowledge (historic dates, world leaders and more) represents the world as it was when they were trained. This is clearly (and somewhat embarrassingly) demonstrated when three of the world’s leading language models (including our own Jurassic-1) still claim Donald Trump is the US president more than a year after Joe Biden was sworn into office.
Jurassic-X solves this problem by simply plugging into resources such as Wikidata, providing it with continuous access to up-to-date knowledge. This opens up a new avenue for AI-assisted text generation on current affairs.

Who is the president of the United States?

T0
Donald Trump
GPT-3
Donald Trump
Jurassic-1
Donald Trump
Google
Joe Biden
Jurassic-X
Joe Biden is the
46th and current
president
Jurassic-X can assist in text generation on up-to-date events by combining a powerful language model with access to Wikidata

Performing math operations

A 6 year old child learns math from rules, not only by memorizing examples. In contrast, language models are designed to learn from examples, and consequently are able to solve very basic math like 1-, 2-, and possibly 3- digit addition, but struggle with anything more complex. With increased training time, better data and larger models, the performance will improve, but will not reach the robustness of an HP calculator from the 1970s. Jurassic-X takes a different approach and calls upon a calculator whenever a math problem is identified by the router. The problem can be phrased in natural language and is converted by the language model to the format required by the calculator (numbers and math operations). The computation is performed and the answer is converted back into free language.
Importantly (see example below) the process is made transparent to the user by revealing the computation performed, thus increasing the trust in the system. In contrast, language models provide answers which might seem reasonable, but are wrong, making them impractical to use.

The company had 655400 shares which they divided equally among 94 employees. How many did each employee get?

T0
94 employees.
GPT-3
Each employee got 7000 stocks
Jurassic-1
1.5
Google
(No answer provided)
Jurassic-X
6972.3
X= 655400/94
Jurassic-X can answer non-trivial math operations which are phrased in natural language, made possible by the combination of a language model and a calculator

Compositionality

Solving simple questions might require multiple steps, for example - “Do more people live in Tel Aviv or in Berlin?” requires answering: i. What is the population of Tel-Aviv? ii. What is the population of Berlin? iii. Which is larger? This is a highly non-trivial process for a language model, and language models fail to answer this question (see example). Moreover, the user can’t know the process leading to the answers, hence is unable to trust them. Jurassic-X can decompose such problems into the basic questions, route each to the relevant expert, and put together an answer in free language. Importantly, Jurassic-X not only provides the correct answer but also displays the steps taken to reach it, increasing the trust in the system.

Do more people live in Tel Aviv or in Berlin?

T0
Berlin
GPT-3
There are more people living in Tel Aviv than in Berlin.
Jurassic-1
Berlin and Tel Aviv are roughly the same size
Google
(First hit is a comparison between Tel Aviv and Berlin)
Jurassic-X
More people live in Berlin than in Tel-Aviv

[‘Return population of Tel Aviv’; Return population of Berlin’; Return which is bigger between #1 and #2’]
Step 1: Population of Tel Aviv. Result - 451523.
Step 1: Population of Berlin. Result - 3664088.
Step 3: Which is bigger,  #1 or #2. Result - Berlin.

Jurassic-X breaks down compositional questions, answers the basic sub-questions, and puts together the answer. Importantly, this process is transparent to the user greatly increasing the trust in the system

Dynamic information (like weather and currency exchange rates)

Certain types of information change continuously - weather, currency exchange rates, share values and more. Such information will never be captured by language models, yet can easily be handled by Jurassic-X by integrating it with a reliable source of information. We performed a proof-of-concept on two such features - weather and currency exchange rates, and the design enables quick integration with more sources to solve your use-case.
Weather - a loosely phrased question about the weather elicits an answer from all language models, where language models always return the same answer, regardless of when the question was asked (funny, right?), while Jurassic-X provides an answer based on the actual weather prediction.

I’m going to be in New-York in 3 days. Should I pack my umbrella?

T0
No
GPT-3
Yes, you should pack your umbrella.
Jurassic-1
Yes, you should. The weather forecast is rain.
Google
(Links to weather websites)
Jurassic-X
Yes, you should pack your umbrella, because in New York in 3 days there will be broken clouds and the temperature will be -2 degrees.

Currency exchange rates change much faster than weather predictions, yet the Jurassic-X concept - a language model connected to a reliable source of information - easily solves this problem as well.

How much Moroccan money will I get for 100 bucks?

T0
125 dirhams
GPT-3
100 moroccan dirhams is about 27$.
Jurassic-1
How much is 100 dollars in moroccan money?
Google
100 dirhams = 10.75 dollars
Jurassic-X
100 USD = 934.003 MAD
Jurassic-X combines a language model with access to APIs with continuously changing information. This is demonstrated for weather forecasts and currency exchange rates, and can easily be extended to other information sources

Transparency and trust

Transparency is a critical element that is lacking in language models, preventing a much wider adoption of these models. This lack of transparency is demonstrated by the answers to the question - “Was Clinton ever elected as president of the United States?”. The answer, of course, depends on which Clinton you have in mind, which is only made clear by Jurassic-X that has a component for disambiguation. More examples of Jurassic-X’s transparency were demonstrated above - displaying the math operation performed to the user, and the answer to the simple sub-questions in the multi-step setting.

Was Clinton ever elected president of the United States?

T0
Yes
GPT-3
No, Clinton was never elected as president of the United States.
Jurassic-1
No
Google
Clinton was elected president in the 1992 presidential elections…
Jurassic-X
Bill Clinton was elected president.
Jurassic-X is designed to be more transparent by displaying which expert answered which part of the question, and by presenting the intermediate steps taken and not just the black-box response

Your Turn

That's it, you get the picture. The use cases above give you a sense for some things you could do with Jurassic-X, but now it's your turn. A MRKL system such as Jurassic-X is as flexible as your imagination. What do you want to accomplish? Contact us for early access

Contact us below and we will get back to you shortly.

Thank you!

Your submission has been received!
Oops! Something went wrong while submitting the form.