Presented by AWS
Applications are now open for the AWS Generative AI Accelerator (GAIA), an 8-week global program designed to fast-track the most promising generative AI startups. The deadline to apply is July 10. Now in its third year, GAIA is part of AWS’s broader commitment to advance generative AI innovation worldwide by providing early-stage startups with credits, mentorship, and access to the top technologies in the AWS Cloud driving this space forward.
Forty visionary late-seed to Series A companies that are pushing the boundaries of gen AI will be selected and supported by experts from AWS, and provided with up to $1 million in AWS Promotional Credits. The program not only supports technical advancement — including model training, system optimization, and launch readiness — but also opens the door to marketplace entry, strategic partnerships, and scaled distribution with tailored go-to-market strategies and support.
“By combining AWS’s industry-leading global cloud infrastructure and AWS’s deep technical capabilities, with our high-value partner network and tailored go-to-market support, this accelerator helps startups accelerate product development, refine architectures and unlock new market opportunities,” said Jon Jones, VP and Global Head of Startups at AWS. “It’s a focused, results-driven environment designed to help the next generation of AI companies scale smarter, move faster, and build what’s next.”
How the 2024 cohort is changing the world
GAIA 2024 attracted more than 4,700 applications worldwide, bringing together a diverse cohort of startups. Here’s a look at some of the most successful participants of 2024.
Phagos
Phagos is on a mission to fight the growing threat of antibiotic-resistant bacteria. Antibiotic resistance to the bacteria that cause some of the world’s most dangerous infectious diseases is rising: it’s one of the biggest threats to global health. By 2050, these bacteria are predicted to be more deadly than cancer.
Phagos aims to change that with ever-evolving, bacteriophage-based medicines that fight bacterial infections. It uses machine learning to tap into the untapped potential of genomics and the natural coevolution of bacteria and bacteriophages.
Bacteriophages are microscopic, self-replicating predators of bacteria. Unlike antibiotics, which kill any bacteria they encounter, bacteriophages can specifically target and kill the bacteria they can infect. These treatments offer a dynamic solution, constantly evolving to keep pace with bacterial mutations over time, based on collected data and phage samples, in sharp contrast to antibiotics, which become less and less effective over time.
“We use Gen AI algorithms to be able to predict which phages will be able to infect it and then eradicate it,” says Adèle James, co-founder and CTO. “We also have technical support and architectural support from AWS that help us plan as well the next months and years of the company to develop what we want, what we need.”
They’re currently focusing on animal health as the area where they could make the most significant and immediate impact. Thanks to AWS, “the fact that we might actually be able to replace antibiotics sounded crazy, and now it sounds less and less crazy every day,” says Alexandros Pantalis, co-founder CEO.
AudioShake
AudioShake‘s AI-powered sound stem separation technology has the potential to revolutionize how audio is handled in industries across tech, media, music, sports, and beyond — so it’s no surprise they walked away the winner of AWS re:Invent’s Unicorn Tank Startup Competition. The win highlights how fast audio is becoming central as an information source in today’s tech landscape.
“You can make sound more customizable for people, more understandable, intelligible, accessible,” says Jessica Powell, co-founder and CEO. “You can open up new creative workflows or new entertainment experiences. We’ve had a really good experience with AWS so far and I think we’re really just at the start of what is possible.”
Automated speech recognition (ASR), AI, and machine learning enable greater access to captioning and localization services for diverse content types, and intelligent automation offers possibilities to expand inclusion and access to content. With targeted improvements to speech recognition and multi-track audio support, AudioShake gives creators nuanced localization abilities that retain artistic intent, improves audio processing for ASR, captioning, dubbing services, and more.
AI Hay
AI Hay is a knowledge discovery platform that seamlessly integrates conversational search, social, and AI to transform how users across Asia access collective intelligence, with a keen focus on localization. They’ve started with Vietnam.
“The problem that we’re trying to solve with AWS is quite simple but also quite big,” says Hiep Nguyen, COO. “When everyone has the ability to generate content with AI, that means the volume will increase exponentially. When there’s too much content, the way we search needs to change.”
They’re using AI to give users direct, valuable, and quick answers to questions, while building in avenues to discover new and different human perspectives, places, and things. AI Hay’s conversational UI lets users ask questions in real language, and can understand the intent and the cultural and local context to serve up the most accurate answer possible.
“We think there’s not enough generative AI consumer products that are serving low- to medium-resource languages,” Nguyen says. “And that’s what we want to do with AI Hay: We want to bring the best technology and democratize it to every Vietnamese person.”
HeroGuest
HeroGuest is a digital academy that trains front-line workers, focusing on the restaurant and hospitality industries — including hotels and restaurants — as well as retail shops of every size. Their goal is to give frontline workers the knowledge they need to get up and running fast, and continue to evolve in their roles. The stumbling block for most businesses is that training can be long, tedious, and sometimes inconsistent. HeroGuest is changing that.
“We take all that information, and through generative AI, turn it into bite-sized lessons that can be delivered on a mobile device, and that just makes training and development fun, engaging, and efficient,” says Gabriel Garcia, managing partner and CEO.
Learning is gamified to engage and delight users, and managers can track the progress of each of their employees as they move through lessons and earn diplomas.
“Leveraging the journey of AI with AWS, we believe that there is this huge opportunity to personalize learning, and that also reflects on career progression, higher compensation through higher tips,” Garcia says. “That’s exactly what we want — not only to solve business problems for our customers, but to transform the lives of the frontline workers.”
Unravel Carbon
Unravel Carbon aims to accelerate the world’s progress to a zero-carbon economy by helping enterprises measure, reduce, and report on their carbon emissions. Their mission is to integrate climate-driven decisions into global business practices, and to leave a habitable world for future generations.
They’re combining AI-powered agents with an enterprise climate platform to help organizations embed sustainability into decision-making, scale their impact, and drive real progress. Its AI-powered decarbonization platform transforms semi-structured customer data into actionable emissions insights, peer benchmarking reports, and regulatory disclosures.
“For startups navigating today’s competitive AI landscape, this accelerator program delivers significant value through technical expertise for implementing generative AI solutions, business mentorship for developing effective go-to-market strategies, and vital AWS ecosystem connections,” says Vidur Puliani, director, machine learning, at Unravel Carbon.
“The combination of targeted learning, expert mentorship, and high-profile showcase opportunities creates a structured pathway for young companies to achieve both technical excellence and market differentiation simultaneously.”
PhysicsX
Many of the world’s most pressing challenges are engineering problems, and PhysicsX is using AI to solve them. Numerical simulations have replaced costly physical prototyping in engineering design, manufacturing, and the operation of complex products and processes. But current simulation methods can still be complex, limited, and too slow and resource-intensive for automated optimization over large design spaces.
PhysicsX develops AI-driven, real-time multi-physics simulations, enabling faster iterations and automated optimization, and exploring designs previously out of reach. The platform integrates all stages of the product lifecycle to drive innovation from concept to in-service operations. They joined the AWS accelerator program to fully harness AWS’s cloud capabilities and directly engage the stakeholders shaping the future of industry.
“As high as our expectations were going into the program, they were on the whole exceeded,” Jacomo Corbo, CEO & co-founder, PhysicsX, says. “This program offers a unique blend of technical depth and commercial reach. As a founder, you gain access to an ecosystem that understands both the demands of building cutting-edge AI systems and the importance of aligning those systems with real-world customer needs. The experience was instrumental in accelerating both our platform capabilities and our market adoption, and significantly amplifying our reach and relevance.”
Who should apply
Applications are evaluated based on a variety of factors, including founding team experience, clear industry use cases and target market, product maturity, evidence of traction, team composition, and breadth of generative AI skills (e.g., ML engineers, data scientists).
AWS is especially interested in applicants who are exploring fine-tuning open-source models, pre-training foundation models (FM), or providing critical services in the emerging generative AI tech stack that surround the FM with data preparation, model monitoring, infrastructure provisioning, and more. Companies led by diverse founding teams are especially encouraged to apply.
GAIA welcomes startups from across industries and geographies. The 2025 cohort will include companies from North America, EMEA, APJ, and LATAM.
Apply now!
The AWS Generative AI Accelerator delivers unmatched benefits designed to accelerate both technical growth and market traction. Accepted startups receive:
- Up to $1 million in AWS Promotional Credits to build, train, and launch AI solutions
- Personalized mentorship from AWS experts and industry leaders
- Access to a curated network of partners and investors
- Hands-on support to accelerate product development and go-to-market
- Exclusive experiences and complimentary tickets for AWS re:Invent
AWS Generative AI Accelerator Program details:
- Application period: June 10 to July 10
- Cohort selected by mid-September
- Program kicks off October 13 and continues for 8 weeks, culminating at AWS re:Invent, Dec 1 – 4, 2025
Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact
