In 2019, three eager innovators met at the Believe in Music conference in Anaheim, Los Angeles. Peder and Erling, working for Soundly, had their initial encounter with Stian from Acon Digital at the conference where their conversation quickly revealed a mutual fascination with audio processing and machine learning, sparking the beginning of collaborative brainstorming sessions that eventually sparked the idea behind HANCE.ai.
The collaboration between Acon Digital and Soundly began with Acon Digital recognizing the need for additional data to enhance their already exceptional algorithms. Fortunately, Soundly possessed a vast sound library, making them the ideal data source. This symbiotic alliance allowed both companies to elevate their products, delivering superior solutions to their user base.
Driven by their shared passion and expertise, the founders identified a burgeoning market for real-time and lightweight audio enhancement algorithms. Thus, HANCE.ai emerged as a company with a clear mission – to create an incredibly lightweight and a CPU-efficient audio engine, capable of adapting to diverse audio-related challenges across a multitude of industries.
We interviewed Erling Hoff about their startup journey and to underscore the significance of academic collaboration in shaping their product.
From idea to market - What lessons have you learned and what advice can you provide future novice AI entrepreneurs?
Transitioning from a concept to a market-ready product is a complex endeavor, but it's one that can be navigated successfully with the right approach. For us at HANCE.ai, the starting point was zeroing in on a niche—audio enhancement—where we knew we could make a meaningful impact.
Developing our algorithms in-house was a calculated move that gave us complete control over our technology. This allowed us to create algorithms that are not just efficient but also highly customizable, setting us apart in the audio enhancement market.
Understanding where to focus our energy was another key lesson. While our expertise is in audio processing and machine learning, we also knew we couldn't neglect other important aspects like user experience, marketing, and building partnerships. Balancing these elements was crucial in developing a product that is not just technologically sound but also user-centric and market-ready.
Reflecting on our journey, we've learned the importance of having a clear focus, the benefits of in-house development, and the need to balance technical expertise with other business essentials. These lessons have informed our product development strategy and will continue to guide our efforts in refining and expanding our technology.
NORA.startup is a community for research based innovation. In what way has HANCE.ai collaborated with academia, and what role has this collaboration had in shaping your product?
Collaboration with academic institutions has played a significant role in shaping Hance.ai. Our algorithms are designed using a blend of deep learning and evolutionary computing, which are fields that have seen remarkable advancements in computer science and artificial intelligence. This isn't just theoretical for us; we have a practical engagement with academia. Two master's students are part of our team, actively contributing to our research and development projects. Beyond that, we make it a point to stay abreast of the latest research findings in AI and related fields. This ensures that our work remains at the cutting edge, allowing us to continually refine our technology based on the most current insights.
In the future, HANCE.ai is committed to deepening its relationship with academic institutions. We recognize the invaluable contributions that academia can make to our field, particularly in the realms of research and technological innovation. As we continue to grow, we seek to establish more formal partnerships with universities and research centers. These collaborations could range from joint research projects to developing specialized courses and workshops focused on audio enhancement and AI. Our aim is to create a mutually beneficial ecosystem where academic research can inform and elevate our technological solutions while our real-world applications can offer practical challenges and data sets for academic study. This synergy will help us refine and expand our technology and contribute to the broader scientific community, fostering innovation in audio enhancement and artificial intelligence.
What are some new ideas you are currently working on, and what can we expect from Hance.ai in the future?
At HANCE.ai, innovation is part of our DNA. We're always evolving, especially when it comes to real-time audio enhancement. Our ongoing projects include refining our current algorithms, creating new ones for various audio challenges, and broadening the scope of devices and platforms compatible with our technology.
One exciting avenue we're exploring is generative audio models. These models can produce new audio data, opening up possibilities from music synthesis to speech generation. Leveraging deep learning, these models can understand intricate audio patterns and generate sound that closely mimics the original structure and style.
Another focus is on stem separation, where we dissect a full audio mix into its individual components or instrument groups. Given the intricate interplay between instruments, it’s a complex task, but our algorithms are designed to handle it with precision.
We're also enhancing dialogue clarity in audio mixes with our ’Voice Boost' feature. This automatically elevates the dialogue volume in films, making it easier to hear against background noise or for those with hearing challenges.
Lastly, we're committed to making our algorithms more efficient and trainable. This involves optimizing the code for faster processing and more accurate results, ensuring that our solutions are effective and groundbreaking.
Last but not least: how can NORA.startup support you and your team in future endeavors?
Partnering with NORA offers multiple advantages for Hance.ai. One of the most significant is tapping into NORA's extensive network of educational and research institutions specializing in AI and machine learning. This connection could provide us with fresh perspectives and practical resources to enhance our technology further. Additionally, NORA's events and annual conferences serve as valuable platforms for us to engage with other innovators in the field. These interactions can potentially lead to new collaborations or partnerships that could fuel our growth.
NORA's dedication to innovation aligns well with our own commitment to developing cutting-edge audio solutions. This synergy could help us leverage NORA's resources to accelerate our own innovative efforts. By collaborating with NORA, we can stay abreast of the latest AI and machine learning advancements, ensuring that our technology remains state-of-the-art. Lastly, NORA's research exchange programs and educational initiatives offer our team opportunities for skill development, which would further contribute to the refinement of our technology.
HANCE.ai is getting ready to launch the HANCE Audio Engine on the 2nd of November, 17:00 at Sagveien 23A, Oslo! We congratulate them on their success and hope that many of you attend the launch to watch them demo their product and explore ways to collaborate!
For more information about HANCE.ai, click here or contact Erling Hoff (erling@hance.ai).
