PhotoProAI is an AI-powered tool that can help you create professional portraits from just one photo. It uses AI to automatically improve the lighting, composition, and overall quality of your photo.

To use PhotoProAI, simply upload your photo and select the desired level of enhancement. PhotoProAI will then automatically enhance your photo and provide you with a preview of the results. You can then download the enhanced photo or make further edits using the built-in photo editor.


PhotoProAI is a great tool for anyone who needs to create professional portraits, such as business professionals, freelancers, and social media influencers. It is also a great tool for anyone who wants to improve the quality of their selfies or other personal photos.

Chris: Hi there, please introduce yourself.

Jacky: Hi there, I’m Jacky Fan. And I’m the founder of PhotoProAI. I used to work for Tencent and Bytedance for a few years.I am mainly responsible for R&D related work. I have a co-founder by the name of Wouter Wink who previously worked at Google for many years on the marketing side.

Chris: What was the inspiration behind creating PhotoProAI, and what specific needs or challenges in the field of photo editing did you aim to address with this platform?

Jacky: I have a lot of hands-on experience with AI, and integrated machine learning in consumer products in various ways over the past years. So I wasn’t too surprised to see that the new wave of image generation tools like Dall-E and Stable Diffusion quickly gained a lot of popularity. What did surprise me though was that many companies that jumped on the AI image generation bandwagon hadn’t really fine-tuned their models well, and needed the user to upload dozens, sometimes even 20 or more pictures to generate realistic looking pictures. I challenged myself to build a better model. What I thought was going to be a weekend endeavor quickly became an obsession, but I’m happy to see that all the time I invested was not for nothing.

Chris: Portrait enhancement is a critical feature, especially for users looking to improve the quality of their portraits. Could you provide more insight into the technology and algorithms used by PhotoProAI to enhance facial features, skin tone, and lighting in photos?

Jacky: We use a combination of open source AI image generation libraries, and proprietary tools and methods that we developed over the course of the past year or so. Our goal was to reduce the amount of pictures users had to upload to train their personalized model, and to do that we had to come up with a few clever ways to simplify the model. For example, when it comes to skin tone, we modeled it to make certain assumptions so it requires fewer inputs. That works well on any skin tone you would see out there in the world. But it doesn’t work if someone has say blue or green skin, because the model is programmed to assume that is impossible. In practice, shortcuts like these work really well.

Chris: Photo restoration is often a labor-intensive process. How does PhotoProAI automate the restoration of old and damaged photos, and can you provide examples of the kind of improvements users can expect?

Jacky: This is actually not incredibly complicated. To our model, an input picture that is grainy, old and damaged is very similar to a selfie that is taken in poor lighting, with hardly visible eyes, and hair that is covered by a hat. In both cases, the model needs to figure things out using missing information. So from an engineering perspective, it’s just not that different a challenge.

Chris: Photo effects are a fun way to personalize photos. What types of filters, frames, and stickers are available in PhotoProAI, and how can users apply them to enhance their photos?

Jacky: Yes, we’ve offered various styles from the start, for example an office setting, school setting, or outdoor setting. We also offer a few original options such as Superman and Barbie. And we allow users to tweak their hair style and eye color. Though it’s a lot of fun to play with, in practice most of our users seek out the fairly straightforward styles like the office setting I mentioned earlier. My assumption is that in the end, what people really need this for is for LinkedIn, their website, or other use cases that require a professional picture.

Chris: Accessibility and ease of use are crucial for photo editing tools. How does PhotoProAI ensure that users of all levels of experience can easily navigate and utilize its features?

Jacky: At PhotoProAI, accessibility and ease of use are built in from the start. In fact, that’s what the product is about. The value proposition is exactly that: anyone can generate a great looking picture. It was never our intent to add a lot of bells and whistles. Rather, we’d like to improve the AI model, making bells and whistles completely unnecessary for a great user experience!

Chris: Many users may be concerned about the privacy and security of their photos. How does PhotoProAI handle user data, especially when dealing with sensitive or personal images?

Jacky: This is a very valid concern. Of course with my background, I’m very well aware that privacy and security are top of mind with users everywhere, and fortunately I have a lot of experience building for scale, and building with global privacy regulations in mind. People will simply not use our product if they don’t feel we have the appropriate measures in place to ensure their pictures stay theirs, and just theirs. We do a couple of things to ensure this. For example, for our front-end development, we use tried and tested frameworks, and don’t try to reinvent the wheel. For our backend, we use cloud storage hosted at some of the better known cloud providers.

Chris: In a competitive photo editing market, how does PhotoProAI plan to stay innovative and up-to-date with the latest trends and demands in photo editing and enhancement?

Jacky: We decided that our competitive edge is going to be that we’re going to do just one thing and do it really really well: create realistic-looking pictures that people can use on their website, social profiles and so on. Other services spread out and offer all sorts of gimmicky add-ons. At least for the foreseeable future, we do plan to spread ourselves thin. It takes a lot of time and skill to keep up with the latest in AI assisted image generation, and we need to be at the cutting edge if we want to keep happy customers.

Chris: Looking ahead, are there any exciting updates or new features in the pipeline for PhotoProAI that users can anticipate in the near future?

Jacky: Well we just released a completely new model, that allows the user to upload more than just 1 selfie so that we can better train their personalized model. Our first product, that is still available, required just a single selfie, and generated a new picture almost instantly. It’s great for when you’re in a rush, but it doesn’t provide a ton of variety. So with this new model, we allow users to upload more pictures, and wait for up to 24 hours to receive 50 different images. In the near future we intend to keep both models available on the website. One is fast but inflexible, the other slower but with a lot of variation. Based on user feedback, we’ll decide how to continue to develop these in the future.

Chris: Thanks for being with me, any last words? Where can our readers follow you?

Jacky: First off – let me say thank you for taking the time to read this. As a developer I can really geek out on the science behind this, but I realize all too well that for most people, all they care about is a product that works really well. I also want to say that we really value user feedback. It’s helped me in the first few weeks of development, but also more recently when we decided to built this entirely new model – based on feedback from users that they wanted even more variety and didn’t care if it took a little longer. And yes we try to keep our LinkedIn up-to-date. Thanks!