- These 12 generative AI startups can build creative, power interactive video, and more.
- Investors include the music artist John Legend and the VC arms of Google and L’Oréal.
- Some of these under-the-radar companies haven’t even officially been announced yet.
When generative AI first became popular nearly a year ago, it was obvious that it would have a significant impact on the advertising industry. However, as questions about potential copyright infringement, hallucinations, and the quality of the technology’s output arose, it became unclear how the technology could be used effectively.
However, a number of adtech and martech companies have emerged in the last year, each employing some form of generative AI technology to do things as diverse as making it easy for marketers to create high-quality brand assets, inserting products into videos, and creating synthetic data models to improve ad targeting.
Insider sifted through hundreds of submissions, interviewed industry experts, and relied on its own reporting to compile a list of 12 promising generative AI-powered startups that are transforming the advertising industry.
All of the businesses highlighted here are startups that rely on generative AI technologies to power their core solutions. While a few have received tens or hundreds of millions of dollars in investment, the vast majority are still in the early stages of development and have received only seed or pre-seed funding. Some have yet to emerge from hiding.
In alphabetical order, here are 12 generative AI startups poised to revolutionize advertising and marketing in 2024 and beyond.
Funding: Paladin Capital will lead a $5 million seed round in 2021, with participation from Reinventure Capital, In/Visible Ventures, and music artist John Legend.
Number of employees: 22
What it does: Adventr, based in New York, is essentially a video Siri. Its “SmartListen” feature allows users to interact with videos in real time by utilizing patented AI-powered voice control.
Devo Harris, the founder of Adventr, was previously a Grammy-winning music producer and composer. executive who has worked with artists such as Kanye West and John Legend. Legend, who also happened to be Harris’s college roommate, is now an Adventr investor.
Harris released a video from the band R! in 2010.ot in Paris for its track “Attack of the 5 ft. Hipster” It received a negative response in the YouTube comments. The following week, Harris shared the same video but added interactive buttons that allowed viewers to select which Scene to watch next. It was a big hit.
“The next thing I know, all these media companies reached out to me and I then spent the last decade diving into video tech,” Harris was quoted by Insider as saying.
Adventr provides cloud-based software that allows anyone to create “smart media,” which allows users to interact with videos to ask questions, buy products, or initiate text messages or phone calls. It also provides an enterprise product, which has been used by clients such as Chevrolet, Disney+, Shopify, AWS, and X — the social media platform formerly known as Twitter — to integrate their databases or apps so that their videos can provide real-time responses to queries.
23andMe used Adventr to allow customers to ask a video of a genetic counselor whether they are pregnant or not. For example, they are more likely to inherit a disease from their parent. They can also use their voice to participate in a DNA quiz gameshow.
Adventr plans to incorporate more large language models and generative AI technology into its voice controls over the next six to twelve months, according to Harris. Users will be able to ask a retailer’s video ad, “I like this jacket, but do you have it in blue?” Harris explained. Adventr also wants to be able to integrate the checkout process so that customers don’t have to leave the video they’re watching to make a purchase.
Funding: Anoki raised a pre-seed round in the first quarter of 2023 but declined to disclose the amount because it is currently raising its next round.
**Number of employees:**30
Anoki personalizes video ads with generative AI. Its technology can detect the context in which ads are displayed and make changes to make the ads more engaging for the intended audience. For example, Anoki’s tech can change a person’s accent or appearance in a video, or make it appear as if he is wearing the jersey of a local sports team. or change the background or the type of product displayed.
Anoki creates ads using a combination of generative AI technologies, including ChatGPT. The final results are approved by Anoki’s human staff.
According to cofounder and CEO Raghu Kodige, Anoki has built out its technology and is working to bring its solution to market, with the goal of launching at CES 2024. Anoki intends to sell its product to agencies as well as brands. and is collaborating with ad-selling media companies to ensure it has access to inventory.
Kodige previously cofounded the television data company. Alphonso was acquired and rebranded as LG Ads by the electronics company LG in 2021. Kodige was in charge of that division until last December.
Chalice Custom Algorithms
Funding: Aperiam Ventures and The Trade Desk’s TD7 Ventures contributed $1.8 million in two pre-seed rounds.
**Number of employees:**20
What it does: Chalice was founded in 2020 by Adam Heimlich, a longtime ad agency executive, and Ali Manning, a former Google and Snap executive.
Chalice is a software company that creates algorithms for advertisers to use in demand-side platforms (the software used to buy digital ads). Chalice’s algorithms have the potential to improve ad targeting. It is comparable to AI. Unlike Google’s Performance Max and Meta’s Advantage Plus, the data that advertisers provide to Chalice is not used to power other companies’ ads across a vast network.
Chalice and its advertiser clients do not have access to the massive amounts of data used by Google and Meta to power their algorithms. This means that prior to the advent of generative AI, Chalice could not assist smaller or mid-sized businesses lacking large data sets, or large advertisers wishing to run smaller campaigns.
Chalice began developing a solution that used generative AI to create synthetic data to power ChatGPT last summer, a few months before OpenAI’s ChatGPT monopolized the newscycle. Clients with small data sets require algorithms. The solution was created from open source code found on the programming website Stack Overflow, and it employed a machine learning algorithm known as a “generative adversarial network.” which is frequently used to create deep forgeries.
“We’re attempting to create a deep fake of the customer’s data that isn’t significantly different from the real thing.” “It’s completely different and almost unrecognizable from the original,” said Tylynn Pettrey, Chalice’s head of data science who spearheaded the company’s generative research. AI is a solution.
This tool has reduced Chalice’s time spent working with clients from three to six months to one to two weeks. It has also assisted Chalice in increasing its client base from five to six a few years ago to around 30 today.
Through the agency Canvas Worldwide, it has clients such as Hershey’s and Hyundai.
Funding: $1 million from LiveRamp Ventures and Aperiam Ventures, among others.
Number of employees: 4
What it does: Adtech entrepreneur Zeev Neumeier founded GraySwan Systems to use generative AI to help marketers spot anomalies, such as when the same ad appears to someone multiple times in a row or when there is an unexplained increase in website traffic that could be due to a successful ad campaign or bot traffic.
GraySwan Systems, according to Neumeier, employs generative AI to examine all aspects of an ad campaign.
It collects data from the demand-side and supply-side platforms used to run programmatic campaigns. The data is then processed by the company’s AI and displayed on a dashboard. According to Neumeier, the dashboard enables marketers to quickly identify ad problems in a matter of minutes.
GraySwan intends to pitch its technology to media companies and publishers first, followed by adtech firms and advertisers. GraySwan Systems currently has one customer using its product and plans to launch it in the first quarter of 2024, according to him.
Neumeier’s previous startup, Cognitive Networks, was acquired by TV manufacturer Vizio and rebranded as Inscape, forming the foundation of the company’s ad business.
Funding: $2.75 million in pre-seed funding has been raised from investors such as adtech executive Brian O’Kelley and MediaLink CEO Michael Kassan.
Number of employees: 19
What it does: Notable cofounders Sebastian Acevedo and Camilo Fosto have created a memorable product. Harvard University and the Massachusetts Institute of Technology (MIT) both have machine learning backgrounds.
Memorable employs artificial intelligence to assist brands such as Unilever, L’Oreal, and NotCo in developing the most effective creative. Its machine learning models predict how likely it is that someone will click on an ad or remember a brand after seeing a social, display, or streaming TV ad.
Memorable models are initially trained on Memorable’s own generative AI tool is used to generate new images for ads based on creative assets for specific brands. Then The startup investigates which creative assets perform best. An ad with a logo in the bottom corner of the screen, for example, may perform better than one with a logo in the center.
For NotCo, a model predicted ad engagement on Meta and then used the generative tool to generate 48 creative assets in five minutes. When compared to human-made creative, generative ads increased engagement by 16% and click-through rate by 32%.
The goal is to reduce the amount of time spent. that advertisers spend on testing creative, according to Acevedo. “If you think of the old ways of optimizing creative, you need humans, you need maybe surveys to pre-test them and so on,” he went on to say.
Brands pay an annual fee to use Memorable’s tools, which varies depending on the amount of advertising they run. Acevedo stated, “spend.” Memorable has also formed alliances with adtech companies such as tvScientific, Vistar Media, and Kargo. Memorable is developing tools to assist brands in predicting the best-performing creative without having to pay for ads to test the performance.
Funding: Individuals contributed $11.4 million in seed funding. Conde Nast Chief Business Officer Craig Kostelic and former Salesforce Chief Scientist Richard Socher are among the investors. Omneky also received funding from venture capital firms such as Softbank’s DeepCore, Village Global, AIX, and Orbit Venture Partners, among others.
Number of employees: 54
What it does: Omneky creates ads using generative AI based on what type of creative will perform best.
To generate marketing material, it employs what founder Hikari Senju refers to as a “fine-tuned Midjourney” that is trained on an advertiser’s existing content. It employs a separate predictive model to guide the development of that content, based on how the design, such as the layout, text size, and image used, will drive clicks or other performance metrics.
Omneky integrates with companies such as Oracle, Snowflake, and Databricks to obtain first-party data from its advertiser clients, and it also integrates with platforms such as Reddit, TikTok, Google, LinkedIn, and The Trade Desk to obtain third-party data. Both distribute the ads and learn how they performed.
Senju told Insider that Omneky has an ad generation tool that all marketers can use, and that it currently generates around $3 million in revenue.
Omneky will release a workflow tool in the coming months, allowing large agencies and enterprises to easily generate thousands of variations of ads and test which ones perform best. This tool is currently in beta testing with a small group of Omneky clients and will be available in the coming months.
“The productivity gain of our tool is such that it’s a thousand X improvement,” Omneky CEO and founder Hikari Senju told Insider. “The The reality of how people make ads is that they still make them one at a time.”
Senju, who founded Omneky in 2018, studied computer science at Harvard, with a focus on AI and machine learning.
Omneky currently has around 100 clients, including Sony Music, Japan Airlines, and StartEngine, a startup investment platform.
Funding: $1 million from irrvrntVC and other investors.
Number of employees: 5
What it does: Steven Liss and Michael Bishop, cofounders of OpenAds, want to create an AI ad network for publishers.
The technology serves advertisements in generative AI chatbots embedded on publisher websites. The goal is to assist publishers in profiting from the surge of AI-generated content. Users enter prompts into OpenAds’ chatbots and are shown ads based on their searches. Publishers will be able to sell the ads in the same way that other native formats do. High-CPM editorial content, brand-safety controls, and contextual targeting based on first-party data.
Bishop previously worked at Oracle-owned Moat, where he developed technology used by publishers to track metrics such as viewability and ad fraud.
“It feels a lot like search advertising, but with a dynamic copywriter,” Liss told me.
The company uses data from 10,000 advertisers to train its AI, and it collaborates with AI and chatbot companies.
This year, OpenAds also took part in AWS’ Generative AI Accelerator program, receiving $300,000 in AWS credit.
Funding: $9 million in Series A funding led by Nadathur Estates, with Spearhead Capital participating.
Number of employees: 77
What it does: In 2019, Quilt.AI released its first generative AI product, which used its platform to understand data and convert that into marketing copy, slogans, and creative.
“We dropped the product on Product Hunt and about nine people signed up,” said Anurag Banerjee, cofounder and CEO of Quilt.AI.
Quilt.AI has since collaborated with major corporations such as Procter & Gamble and De Beers. Quilt.AI is used by Beers and Mastercard to convert big data signals from video, image, and text across the web into meaningful insights about customer and broader human behavior.
Quilt.AI also provides a “marketing AI” that generates image and text content based on those insights. Video will be available next year.
In addition, the company collaborates with nonprofits such as The Gates Foundation and The World Bank to help them understand patterns in areas such as mental health and climate change, and it donates 5% of its revenue to charity each year. month.
Banerjee stated that he wants the company to be integrated into all of the major language models available from big technology. companies. He also wants Quilt.AI to develop 100 “mini foundational models” around categories such as beauty, travel, and football, as well as more nuanced indices around topics such as “British humor.” Quilt.AI also wants its generative AI model to be “fully automated.” “transparent,” so that its clients understand which data points are driving certain actions and what the model’s parameters are.
Funding: In June 2023, Greycroft, United Talent Agency’s VC arm UTA Ventures, and L’Oréal’s VC fund BOLD raised a ****$8 million seed round.
Number of employees: 40
What it does: Rembrand employs artificial intelligence to place animated virtual products in creator videos. A shampoo bottle, for example, may dance across the host’s desk.
“We use AI generative techniques to photo-realistically fuse brands into the content that people really want to watch, without interruption,” he explained Omar Tawakol, Rembrand’s cofounder and CEO.
The placements do not necessitate the presence of a physical product, nor do they necessitate the use of a VFX team to insert a digital object in post-production. Advertisers, on the other hand, can select from a library of animations that determine how the digital product moves through the video.
Rembrand’s AI separates people and moving objects in the video, so a product on a shelf in the background is always behind the host and does not, for example, appear in the foreground. If she moves in front of it, it will obstruct her face. In the video, Rembrand generates a large amount of synthetic data about the environment to determine how the product should interact within that space.
Rembrand works with clients in three ways. First, advertisers can provide Rembrand with their products. creative assets and pay for it to run an ad campaign. Second, advertisers can use Rembrand’s technology to create their own campaign with their preferred creators. Third, Rembrand is attempting to collaborate with movie studios and premium content creators in order for them to use Rembrand’s technology to carry out their own product placement deals.
Rembrand collaborates with approximately 15 creator agencies, including UTA (an investor), to gain access to their full talent rosters, and it has worked with approximately a dozen of them. Tawakol declined to name four global brand campaigns that he oversaw.
Tawakol’s previous two startups, BlueKai and Voicea, were acquired by Oracle and Cisco, respectively.
Funding: $21.5 million from Disruptive AI and PeakBridge, among others.
Number of employees: 75
What it does: Tastewise, based in Israel, assists food and beverage companies in developing new products. Without using consumer focus groups and surveys, you can generate product ideas and marketing campaigns quickly.
Alon Chen, CEO and cofounder of Tastewise and former CMO of Google in Israel and Greece, is a self-taught engineer who came up with the idea of Tastewise after realizing at his workplace that consumers’ tastes are changing faster than mother’s Shabbat dinner New products can be developed by brands.
TasteGPT, a six-year-old startup, has proprietary generative AI tech that giant consumer packaged goods brands like PepsiCo, Kraft Heinz, and Campbell Soup Co. use. Brands use ads, images, packaging design, blog posts, and recipes to market their products.
While Tastewise began by assisting businesses with product development, it has since expanded. Expanding to assist food and beverage brands in reducing advertising costs. Tastewise’s AI, for example, can create a product based on a food term like “hamburgers” and generate marketing for that new product.
Funding: $165 million from investors including Salesforce Ventures, Lightspeed Venture Partners, Madrona Venture Group, Menlo Ventures, and Microsoft. M12, as well as Google’s investment arm GV.
Number of employees: Approximately 75
What it does: Adobe’s former CTO and Chief Product Officer Abhay Parasnis founded Typeface in 2022, and it emerged from stealth in March. Parasnis was also a member of the team that created Microsoft Azure over ten years ago.
Typeface employs generative AI to enable brands and agencies to create content quickly and cheaply. This content can also be customized at a large scale. “The paradox for most companies is they can choose either high-quality content, which is slow and expensive, or fast content which isn’t on brand,” Parasnis told Insider in a telephone interview.
Typeface avoids copyright issues by training on assets owned by its advertiser clients, ensuring that it can accurately capture things like the brand’s voice or the details of its logo or products.
The company accomplishes this through “personalized AI,” which is a combination of large language models from companies such as OpenAI and Google, as well as content injection. and rules tailored to the advertiser. Users can then ask Typeface to generate creative content by using either a prebuilt template or prompts.
According to Parasnis, Typeface is not attempting to replace high-end creative tools such as Photoshop or Canva. Instead, Typeface is intended to assist marketing staff in distributing content as quickly as possible.
LG and Sequoia Consulting Group are among the typeface clients.
Despite its recent launch, the startup has already secured several significant partnerships. For example, it collaborated with Google to launch a campaign creation solution.
It partnered with Microsoft so that people using Teams can build content, such as if staffers want to quickly use its AI to convert a webinar or video into an email campaign or blog post). This capability was revealed in August.
And, in September, Typeface announced a partnership with Salesforce to launch a solution that can create images and drop them into email marketing campaigns.
Funding: $6.9 million in investment from a variety of sources, including Draper Associates, Rise of the Rest, Detroit Venture Partners, North Coast Technology Investors, and Invest Detroit are some of the firms involved.
Number of employees: 27
What it does: Founded in 2017, Waymark’s AI technology extracts data from a company’s database. To create a ready-to-air TV commercial, use the website URL and its social media.
It collects creative assets to build a custom commercial and voiceover script using a proprietary system of AI models, including ChatGPT and other non-language models. Users can then make changes to any of those elements.
Waymark has agreements with major television networks. FOX TV, Gray Television, Morgan Murphy Media, and the Australian media company Nine are among the cable operators. Spectrum Reach, Charter Communications’ ad sales arm, created over 4,000 ads in 2022. According to Waymark, the spots generated $11.5 million in ad revenue from small and medium-sized businesses.
Waymark’s ad creation technology is also integrated into self-service ad platforms like Spectrum Reach, Hulu, and Roku.
Waymark intends to use a more robust and advanced LLM to enhance its operations in 2024.