Presented by:

Topic #1

Retell AI (trends)

Retell AI provides a platform for building and managing AI voice agents.

These AI voice agents can answer customer phone calls or make outbound phone calls all while using the business’ existing phone number.

The voice agents can be programmed to answer questions, schedule appointments via direct integration with Cal.com, or transfer the call directly to a human.

Developers can choose voices from ElevenLabs, PlayHT, or OpenAI, and program them to operate in 18+ languages.

Retell AI leaders say the latency of their voice agents is what sets them apart from competitors.

With response times averaging 500 milliseconds, the bots respond just as fast as human operators. The bots can also handle interruptions.

Company leaders report that they’ve worked with 3K clients so far.

In April, Retell AI hit $60M in annualized revenue. 

What's Next

Retell AI is part of the AI Customer Service meta trend.

Search growth in “AI customer service” has skyrocketed over 3200% in the past two years.

Looking specifically at AI in the contact center, predictions show the overall market value could hit $3B by 2028.

The CEO of Zendesk has said that 100% of customer service interactions will utilize AI in the near future.

Here are three startups developing AI tools for customer service:

Slang AI is an AI-powered tool specifically designed to answer phone calls at restaurants. Their voice agents can answer questions and book reservations. Company data shows that restaurants see a 20x ROI with Slang.

Level AI provides an end-to-end AI-powered platform for call centers. The tool monitors interactions, creates coaching programs for agents, and provides an AgentGPT tool for use during customer calls. The startup has more than $73M in funding.

Decagon AI provides a conversational AI tool that works via chat, email, or voice. Users can format Agent Operating Procedures to teach the AI how to respond to unpredictable customer questions. Earlier this year, Decagon raised $250M in a Series D funding round.

Attend

Set a winning strategy for 2027

Senior practitioners from OpenAI, Getty Images, Johnson & Johnson, the Financial Times, Visa, Wix, and more are sharing what they've built, tested, and shipped. So you can do the same.

Find the keynotes that matter most to your 2027 strategy:

• Your next customer might not be human. Is your brand ready?
• The 7 cardinal sins of AI: Stop experimenting, start shipping
• Human stories, machine speed: The future of video content
• How the best B2B brands win before the buyer knows they're buying

Get tickets for 30% off (just £175) until June 23.
 

Claim 30% off 👉

Topic #2

Muon Space (trends)

Muon Space is a startup that designs, builds, and operates low Earth orbit satellite constellations.

These satellites use sensors to track events on Earth, such as weather patterns, ocean winds, and land surface temperatures.

So far, Muon Space has launched seven satellites.

The most recent satellite launches were commercial radio frequency sensing and analytics satellites built for Sierra Nevada Corporation.

With $100M worth of contracts signed recently, the startup says many more satellites are coming.

At least a few of those satellites will be for the U.S. Air Force and Space Force.

These new satellites will replace the aging Defense Meteorological Satellite Program.

By the end of 2027, Muon Space leaders say they’ll be producing 100 satellites annually.

The startup closed a $146M Series B funding round just last year. That’s in addition to a $57M Series A the year prior.

What's Next

Muon Space is part of the Satellite Startups meta trend.

Annual funding for spacetech startups has topped $6B for 2023 and 2024, and last year, funding reached an all time high of $12.4B.

But the satellite startup space is standing out from the others.

A Goldman Sachs analysis reported that the market for low Earth orbit satellites could grow from its current value of $15B to as much as $457B by 2035.

Here are four startups driving that growth:

GalaxEye is building multi-sensor satellites with numerous applications like tracking troop movements, tracking crop growth, and monitoring water quality. The startup has raised more than $18M.

OroraTech has a constellation of heat-sensing satellites that can detect wildfires as small as 10 x 10 meters. They also use AI to remove false positives and predict where the fire will spread. OroraTech has raised about $70M.

CesiumAstro is focused on communications systems. Their satellites include phased array antennas for high-speed, multi-beam communications in space. The startup has raised more than $655M.

Slingshot Aerospace provides space object tracking. Satellite operators use the platform to avoid collisions and plan maneuvers. The startup has raised over $150M in funding.

Topic #3

Data Poisoning (trends)

Data Poisoning is a type of cyberattack that compromises training datasets for AI and ML models.

Cybercriminals may intentionally add false or misleading information, alter the dataset in some way, or delete portions of the dataset.

Compared to ransomware, this is a relatively subtle attack. Users can still access their models and datasets, and so the attack may go unnoticed for a period of time.

Eventually, users will see the models are producing unintended outputs and biased results. It may even result in a backdoor attack.

A data poisoning attack could have devastating results in sectors like healthcare, finance, and government.

For example, a poisoned healthcare AI system may switch a patient’s blood type or recommend an incorrect diagnosis.

One study specifically focused on medical LLMs found that it doesn’t take much data manipulation to cause problems. Replacing just 0.001% of training tokens with false information is enough to produce serious medical errors. 

What's Next

Data Poisoning is part of the AI Cybercriminals meta trend.

Search volume for “AI cyberattack” has quadrupled in just these past few months.

There are countless ways AI can be used in cyberattacks.

For instance, AI models can be prompted to write code for malware.

AI can also speed up the process of finding vulnerabilities in software and updates.

But today, hackers most commonly use AI to create personalized and hyper-realistic phishing emails, deepfakes, and voice clones.

Earlier this year, the FBI warned that hackers were using AI-generated voice messages to impersonate government officials.

In the future, cybercriminals could create entire AI ecosystems to automate hacking.

Topic #4

Cayenne Pepper Supplement (trends)

A Cayenne Pepper Supplement is a dietary supplement made from the cayenne chili pepper.

These peppers contain capsaicin. As well as producing the spicy flavor, this is linked to a number of potential health benefits, including pain relief, better digestion, and antioxidant qualities.

Cayenne pepper is also used as a weight loss supplement. One study found that it reduced appetite by 10%.

The supplement can be taken in various forms. Cayenne pepper drops have 3700% search growth in the past 2 years.

One popular cayenne pepper solution on Amazon has made 16.5K+ sales in the past month, generating revenue of approximately $520K.

What's Next

Cayenne Pepper Supplements are part of the Capsaicin Health meta trend.

Over the course of a 3-month study, capsaicin “significantly reduced” heart disease risk factors among adults with low levels of HDL cholesterol (good cholesterol).

And when it comes to pain relief, another study found a 37% mean reduction in self-reported pain scores after a year of applying capsaicin patches.

The capsaicin market is worth an estimated $308M. That could rise to nearly $460M by 2032.

Capsaicin gel is a muscle and joint soothing cream. One brand, Capsazin, is ranked 23rd in Amazon’s arthritis pain relief category.

Qutenza is a prescription-strength capsaicin patch for diabetic or post-shingles nerve pain. Uptake in the US has more than doubled in the past 2 years.

“Exploding Topics Pro helps me identify trends that no one else is highlighting”