Episode 104: AI Revenue Optimization Secrets with Simeon Lukov Dynamic Pricing

Recently Shopify connected with OpenAI and once you start chatting and searching for a product you may create direct purchase from OpenAI without knowing the merchants or the brand.

That might be a new thing in e-commerce of course with some budget limitations and some human in the loop. That's the next thing in commerce that is coming.

Hi, my name is Demetri Panici and I'm a content creator, agency owner, and AI enthusiast. You're listening to the AI Agents Podcast brought to you by Jotform and featuring our very own CEO and founder, Aytekin Tank.

This is the show where artificial intelligence meets innovation, productivity, and the tools shaping the future of work. Enjoy the show.

Hello and welcome back to another episode of the AI Agents Podcast. In this one, we are featuring Simeon Lukov, the founder and CEO of Dynamic Pricing AI.

How you doing, Simeon?

I'm doing great. How are you doing?

I'm doing awesome and living the dream. That's what I always say. Maybe it's not the truth, we don't know.

But no, in all seriousness, Simeon, thank you so much for making the time today. Hopefully, you're living the dream building out your company right now.

How did you first and foremost get into AI and how did you get into this kind of realm of AI specifically?

Well, I used to be in the e-commerce business back then in 2006 and I saw that I need to do something better in terms of tooling.

From e-commerce, I did my first and second startup and the second one is dynamic pricing for e-commerce. This is how I got into AI.

Also, my background is more technical. I've been CTO and I'm interested in math. That's my education, algorithmic and math, and it's also related to AI in a way.

Yeah, totally fair. What about your company? What is unique? Tell us a little bit about what you guys do and how it stands out from the crowd.

Sure. Dynamic Pricing AI is right now focusing on high volume merchants in the Shopify ecosystem and is bringing new capabilities for price testing, rapid price testing, revenue and profit optimization, and lots of pricing rules and widgets for e-commerce merchants.

Very cool. Tell us a little bit about what that looks like for companies. What do you mean? Just dive into it a little bit more.

Obviously, you have different companies that use your product. Give some cool examples of how it works and how you help people out.

People are usually doing three things. First, they're trying to sell on a high margin. For example, you have new collections of some products and you're not sure what's the exact price the market can pay for that one.

You can try selling jeans at $35, but then you see good demand and try higher price points like $40 or $45. That's the first application and how you get more profit.

The second is to check for discount. Sometimes people undercut by putting too much discount on a product while they don't need it. They might try 15, 20, 25, 30% discount and realize that 15% brings the same amount of orders as 30%, so no need to undercut.

The third application is to test the willingness to pay and the perceived value for a particular product. You test higher and lower prices to see the optimal price positioning for your products.

Interesting. I'm looking at your website now. How do you guys structure pricing? What do they pay for and how have people seen ROI come through?

Some products are just sign up and pay but you don't have to disclose total pricing. How does the model work?

We have a free tier to check all capabilities. Once you see revenue coming, you can switch to a paid plan that is currently $79 per month, mostly bound by the number of orders merchants do per month.

Then you have plans at $449 and $999 which bring more AI models, capabilities, AI recommendations, and some agentic workflows.

I want to talk about agentic workflows. What do those look like for your company? Because we're all about AI agents here.

Agent workflows are the biggest thing. What does that term mean to you, especially at Dynamic Pricing AI?

For us, agentic flow and agents mean something that supports merchants in taking decisions, sensing and analyzing lots of data, especially business metrics and KPIs of their portfolio.

That's easily sensed by models and the agent recommends the best next step for merchants. They have pre-built pricing campaigns, review them, and can launch something pre-built from AI agents for their shop.

What have you guys recently released that you're really excited about product-wise?

We have the agentic workflow and a new type of price testing called switchback, where two prices alternate sequentially. It's counterintuitive but very new in price testing.

Switchback allows CRO guys doing optimizations to see and feel what they are actually doing compared to other tools.

Tell me about the integration piece. You have it listed for Shopify, Magento, Salesforce. How easy is it to set up?

Shopify is our target and integration is quite easy due to extensive documentation and nice APIs. Other ecosystems like Magento, Salesforce cloud commerce, and PrestaShop are less strong but still working.

Integrations are easier with Shopify and a bit more challenging with other platforms due to legacy communication but still functional.

Other integrations are with similar complementary apps popular in some spaces, handled case by case.

What is your favorite thing going on in the company right now? Growth, interaction with clients?

The most exciting thing is lots of installations in the past three months on Shopify marketplace. We track what's going on and have more than 200 installs in that time.

Our task is to convert installs into customers, which is challenging but we have many five-star reviews and paying customers. That excites my team and me.

What in the world of AI outside your company has been top of mind recently?

We started experimenting a lot with Gemini models for our AI agents. These are the most exciting things happening, not only models but infrastructure and frameworks appearing on the market.

I like Google's agentic development kit framework and their great documentation, including a book on designing proper agents. This moves the industry from hype to real applications slowly but surely.

Initially it was hype that you can build an agent with a model, but now you need to manage memory, store sessions, and use racks for different purposes so the agent doesn't spend millions doing a task.

What are your long-term goals and ethos for the company? What do you want companies to get out of this overall?

Our next steps are to try dynamic pricing in other industries like train tickets, sports events, and roadside assistance, while exhausting the e-commerce space.

Recent market changes in taxes, tariffs, and unstable situations make people look for tools that quickly operate prices while maintaining revenue and profit.

Looking at your company, there are many AI tools for retail and e-commerce. Yours stands out. What would you like to add to help companies more? Any features in the works?

We talk about AI recommendations and LLMs helping merchants, but the actual work is done by a framework called multi-arm bandit and contextual bandit for fast reinforcement learning.

LLMs predict tokens and can't predict the best next price, so we have proprietary models for dynamic pricing getting reward signals from merchants in sales and penalties.

Based on rewards and penalties, we train fast reinforcement learning AI and use LLMs on top, not the opposite. Models learn from day one, not using historical data but operating and getting feedback from the market.

Models take different contexts like weekend, weekday, time of day, or competitor status to decide the best price in real time. This makes us different from everyone else.

What industries have latched on to this? E-commerce is broad. Are there specific products commonly sold?

Industries selling not expensive but emotional purchases, products up to $200-$300 that people don't think too much about. They need the product and like the current price regardless of small differences.

For merchants, small price differences matter a lot when selling hundreds or thousands of products, adding nice additional revenue.

Why have you seen any higher ticket pricing products adopt AI in e-commerce? Any other improvements in the space?

There are many tactics on the front end. Users might see widgets deciding to purchase a product. We build widgets next to prices showing trending products with price going up or discounts for out-of-season products.

There might be flash sales for 1 to 5 minutes with mystery discounts personalized for you. Personalization takes into account purchase history and offers special prices for your segment.

I've heard about companies using AI for more personalization on their websites, which is intriguing. Where do you think interesting innovations will come from in e-commerce AI soon?

New generation chatbots will do better and more complex tasks, not only helping purchase but showing product locations and after-sales chats with merchants.

Chats might hook into WhatsApp, email, or other messaging. Another hot space is building product listing pages with automation for images, descriptions, and features.

Back end cloud providers allow easy migration from one platform to another via AI, which is cool for agencies and integrators.

People have done this manually in parts. Chatbots have become universal on websites. What impact do you think they have on customer service for e-commerce stores?

The trend will shift to purchasing directly from chatbots like ChatGPT or Gemini, where you search and get proposals in the chat window.

Shopify connected with OpenAI. You may create direct purchases from OpenAI without knowing merchants or brands. Agent-to-agent purchases might happen, where your health agents restock supplements with budget limits and human in the loop.

That's the next thing in commerce coming.

What do you think these improvements will do to jobs? Will AI make jobs easier or remove people from the workforce?

The hype of AI taking over jobs is declining. Some AI godfathers said agents will take care of many things but that time hasn't come yet.

Many job offers are popping up. People are not in the balloon that AI will replace them soon because hallucinations exist and humans need to check AI performance in industries like legal and marketing.

There might be replacement but it could lead to conflicts. Firing many people could cause crises. People see AI and agents as tools, not replacements.

Considering many companies use OpenAI and Claude models, how do you feel about relying on other companies' components? Everyone is on AWS and now stacking tools.

From a business perspective, locking with AI and LLM vendors might not be healthy. Many new apps appear with no big differentiation except distribution speed.

We prefer investing in research and development of additional models independent of LLMs to operate in dynamic pricing without needing LLMs for core jobs, using reinforcement and machine learning to be independent and have advantage.

Finally, what is your favorite AI tool you use consistently?

From OpenAI, it's their charging and latest models. From Google, Gemini 2.5 models. I like how Google is getting back in the game with very fast models delivering responses quicker than others.

Gemini 2.5 Pro deep research mode is great for setting up research knowledge bases for agents with unlimited deep research. I also love Claude 4.5 Sonnet for writing, the best writing tool I've seen.

If you could let everyone know where to check out your product, please do.

You can check out Dynamic Pricing AI on the Shopify marketplace for optimization.

Thank you so much, Timmy, and I appreciate your time.

Everyone, please leave a like, comment, and subscribe. Leave a review on Apple Podcast and Spotify so everyone can see how great Dynamic Pricing AI is. Thanks for watching. Bye-bye.