The success of any product lies in understanding human needs, and this is the driving force behind Seezar. But how do you create a virtual assistant that feels human, solves real problems, and stands out in a crowded market? To find out, we’re joined by Nader Dagher, Director of Product Solutions and Integrations at Seez. In this episode, we dive into his journey of building Seezar—from tackling the challenges in the automotive industry to developing AI-driven solutions, prioritizing key features, addressing client needs, and shaping Seezar into an indispensable tool for dealerships and OEMs alike.
Part 1 : Bringing Seezar to Life
What motivated the creation of Seezar, and what specific challenges were you aiming to address?
Seez has been at the forefront of AI innovation in the automotive industry since our early days, though it wasn’t recognized as AI back then. In 2016-2017, we pioneered the use of Natural Language Processing and Machine Learning to create an innovative chatbot that negotiated car prices with sellers on behalf of buyers. As we expanded our business—launching a marketplace in Denmark and developing digital e-commerce solutions—we identified a growing need for enhanced customer support. Our experience in both automotive retail and AI technology, combined with the emergence of Large Language Models like GPT, presented a unique opportunity, especially as consumer behavior shifted increasingly online during the post-Covid era.
While many companies rushed to launch AI solutions, we took a different approach. We leveraged our eight years of experience with our original Seez R chatbot in the GCC marketplace to develop something specifically tailored for the automotive industry. Our focus has always been on creating a solution that serves both dealerships and their customers, ensuring the best possible experience for everyone involved.
How did you determine the essential features for Seezar to effectively serve both dealerships and customers?
Early in Seezar’s development, we recognized the importance of catering to both dealerships and their customers. Our primary relationship is with dealerships, but we understood that creating an exceptional experience for their customers was equally crucial.
From the dealership perspective, the challenge was clear: not every dealer has the resources to maintain a 24/7 customer support team. Their primary goal was lead generation – understanding what customers are looking for, guiding them effectively, and collecting information that could help convert inquiries into sales or service appointments.
To meet these needs, we developed a comprehensive tool that provides round-the-clock support. When customers visit a dealership’s website, Seezar is there to help them find vehicles, locate information, and capture their details as qualified leads. All this information flows into a backend system where dealers can track and manage their leads efficiently.
For end users, we wanted to break free from the limitations of traditional decision-tree chatbots. We’ve all experienced the frustration of clicking through predetermined options and waiting for more choices to appear. Instead, we focused on creating a natural, conversational experience where users can freely express their needs – whether they’re looking for an SUV, wanting to book a test drive, or searching for a specific car in black.
Our goal was to make the interaction as human-like as possible, moving away from the rigid, robotic feel of conventional support systems. This approach allows for both simple queries and complex requests to be handled seamlessly, providing users with the freedom and flexibility they deserve.
We refer to Seezar as an AI Virtual Assistant rather than just a typical chatbot. Can you explain to our listeners why we make that distinction?
A “chatbot” is a word that’s widely used and often misused. Unlike traditional chatbots that follow rigid decision trees with predetermined paths and limited options, Seezar functions as a true AI Virtual Assistant that offers a natural, conversational experience. It can understand and respond to free-form questions, analyze information from multiple sources, and provide comprehensive answers.
What sets Seezar apart is its user-centric approach. Instead of immediately pushing for lead generation like many solutions do, we prioritize user comfort and freedom. Users can explore and get information without being forced to provide contact details first. The interface is designed for natural conversation rather than click-through options, and dealers can customize it to their specific needs.
Seezar’s capabilities go beyond simple responses – it aggregates information from various sources to provide holistic answers. This means users don’t have to jump between multiple websites for different pieces of information, whether it’s specifications, reviews, or comparisons. Everything is available in one seamless conversation.
As our CPO Andrew says, it’s like walking into a showroom – no salesperson immediately asks for your name and phone number. Instead, they show you the car, answer your questions, provide an overview, and only ask for contact information if you’re interested. Many dealers invest in making their physical showrooms comfortable with nice couches and coffee machines, but their online presence doesn’t match this experience. Instead, users often encounter an unfriendly bot with an AI-generated avatar and a fake name, immediately demanding personal information.
This creates an uncomfortable solution that most people won’t engage with. They’ll either browse the website without finding what they want or simply leave – creating a significant drop-off point in the customer journey.
Part 2 : Building the Ideal AI Assistant
How do you prioritize features during the product development process, especially when working on something as complex as an AI assistant?
Feature prioritization is a complex process that involves both internal and external factors. Our product team works closely with business stakeholders and management to understand what our clients are looking for, while also conducting market research to identify best practices and use cases.
We maintain a quarterly roadmap that outlines our development plans, prioritizing features based on client expectations and value generation. Features that can deliver quick, significant value – such as improving user retention, lead generation, or conversion rates – typically get top priority.
This process is flexible, though. Like any company, we sometimes encounter situations where we need to adjust our priorities based on new partner requirements or emerging opportunities. We regularly reassess our roadmap to ensure we’re building features that deliver the most value to our clients and their customers.
How important is integration with other systems for AI to deliver real business value?
Integration is a key factor for success in AI implementation because data needs to come from somewhere. There are many moving parts when it comes to a product like this – cars, website data, client data, lead data, CRM, and DMS systems. In fact, about 95% of the time, each of these components is a completely separate solution.
To execute what seems like a simple action – such as generating a lead and having the dealer receive it – requires integration with multiple systems. This process involves knowing which cars are available, triggering emails, and managing customer communications. While it might require 3 to 4 different integrations working together, to the customer it appears seamlessly simple.
We maintain numerous integrations as part of our roadmap, including crucial elements like F&I (financing and insurance) and security systems. These integrations are essential for the buying process – customers want to know car costs, financing options, and monthly payments, all of which require different system integrations.
When prioritizing integrations, we focus on value. We typically start with systems that have the largest market share – for instance, a system used by 70% of the market would take priority. However, we’re also flexible in developing custom integrations for clients with unique systems. Sometimes we encounter clients with unique systems who want to maintain their existing setup. In these cases, we’re happy to develop custom integrations to accommodate their needs.
How does Seezar adapt to the evolving needs of dealerships and customers over time?
Seezar takes a proactive approach to staying ahead of market needs. Rather than just responding to requests, we actively identify opportunities for improvement based on our broad exposure to different dealerships. This unique position allows us to observe best practices across multiple clients and suggest valuable improvements before they’re even requested.
We focus heavily on bringing value through technological advancement. For instance, we’ve identified and helped resolve fundamental digital infrastructure gaps, such as missing sitemaps that are crucial for SEO and marketing effectiveness. As technology evolves, we ensure our clients stay current with these essential tools.
When evaluating new features or integration requests, our decision-making process is value-driven. We assess potential impact through multiple lenses: How many dealers will benefit? Will it open new markets? Will it improve lead generation, customer retention, or inventory showcase efficiency? For example, when a dealer requests integration with a new system, we evaluate the system’s market penetration and partnership potential.
Ultimately, our adaptation strategy combines proactive improvement suggestions, careful evaluation of client requests, and strategic partnership development – all focused on delivering maximum value to our dealers and their customers.
What challenges have you faced in scaling Seezar across different regions, markets, or use cases?
Scaling Seezar across markets offers both opportunities and challenges. Our Large Language Models (LLMs) handle multiple languages effortlessly, allowing quick linguistic adaptation from English to any other language. This flexibility has been crucial for our international expansion.
The main challenges lie in backend operations. Each market requires different system integrations (CRM, dealer management systems), data formats (VIN vs. plate numbers), and compliance standards like GDPR. For example, German dealerships needed special adaptations due to their decentralized specialist structure, while Swiss operations required handling both standard German and Swiss German dialects. These market-specific requirements often necessitate significant customization of our core product.
Security and abuse prevention are paramount in our operations. With SOC 2 certification and robust security measures, we’ve prevented common exploits like prompt injection or misuse as a free GPT service. Unlike basic chatbots that can be manipulated (like the Chevrolet $1 car incident), Seezar maintains strict limitations and stays focused on automotive queries. Our security framework includes advanced prompting techniques and continuous monitoring to ensure system integrity. This comprehensive approach has helped us maintain a secure and reliable service across all markets, building trust with both dealers and end users.
How is Seezar able to accurately understand and provide the right information to every question?
Seezar uses a sophisticated multi-agent system that routes questions to specialized agents for sales, service, parts and accessories, or general information. The system handles both simple and complex queries – from comparing vehicles like BMW 3 Series to C-Class, to answering specific questions about features like cupholders or unique calculations.
For sales inquiries, Seezar excels at understanding complex requirements. Users can provide detailed scenarios, like being an expecting parent needing a safe car with space for outdoor activities, and the system converts these requirements into specific filters.
The technical foundation of Seezar’s knowledge comes from a custom-built scraping solution and direct integration with dealer management systems (DMS) for real-time inventory access.
Additionally, Seezar maintains brand integrity by adapting its communication style for each dealership. When representing a Mercedes dealership, it remains focused on Mercedes products while being factually accurate about features and comparisons.
The system presents results through an intuitive chat interface, displaying vehicle images, specifications, and other details without requiring users to leave the conversation. Users can seamlessly move from browsing to booking test drives.
In essence, Seezar’s accuracy stems from its combination of specialized agents, comprehensive data integration, and brand-appropriate responses while providing detailed information for every query.
Part 3 : Making AI Accessible for Dealerships
How do you believe AI is transforming customer interactions, particularly in the automotive sector?
AI is revolutionizing the automotive industry in three key ways: empowerment, speed, and integration. Let’s explore how these elements are transforming the car-buying experience.
First, AI empowers everyone involved in the automotive sector. While experienced salespeople bring valuable expertise, they can’t instantly access every detail about every vehicle. AI fills this gap by providing immediate, accurate information about specifications, comparisons, and options—empowering both dealers and customers to make informed decisions.
Speed is another game-changer. Traditional car recommendations might take a salesperson 20 minutes to compile. Our AI can match customers with five suitable vehicles in just 7 seconds. This dramatic improvement in response time transforms the customer experience.
Finally, integration ties everything together. By connecting AI with e-commerce capabilities, we’ve created a seamless experience where customers can move from browsing to buying in one conversation. They can evaluate trade-ins, submit bank applications, and complete other transactions—all through a single interface.
Just as the internet didn’t replace professionals but made everyone smarter, AI isn’t replacing human interaction. Instead, it’s enhancing the car-buying process by making it more efficient, informed, and user-friendly.
What misconceptions do businesses often have about implementing AI-powered solutions?
There are two major misconceptions businesses have about implementing AI solutions. The first is about AI’s inherent intelligence. While large language models like GPT have access to vast amounts of general knowledge, they aren’t inherently “smart” without proper setup, prompting, and limitations. Many businesses mistakenly believe that simply integrating an AI API and adding “AI” to their logo makes them an AI company. This oversimplified approach can be potentially disastrous for business operations.
The second misconception is about AI’s knowledge capabilities. Businesses often assume that once they integrate an AI solution, it will automatically know everything about their business. However, AI systems are limited to the data they have access to. For example, if business hours aren’t provided to the system, it won’t be able to answer questions about operating hours accurately.
In reality, successful AI implementation requires proper onboarding, maintenance, and continuous optimization. While initial setup can be quick, it’s not a true “plug-and-play” solution that can perform optimally without proper data input and ongoing refinement. The key to effective AI implementation lies in understanding these limitations and properly configuring the system with relevant data and appropriate constraints.
What are some common challenges dealerships face when trying to adopt AI solutions, and how do we as Seez simplify the process?
At Seez, we’ve made AI implementation remarkably simple for dealerships, regardless of their technological readiness. Understanding that many dealerships still rely on traditional methods like Excel sheets and paper files, we’ve streamlined the onboarding process to be incredibly straightforward. All a dealership needs to provide is their location, operating hours, and website URL – we handle everything else. Our automated setup process includes website scraping, inventory collection, and bot training, culminating in a simple script that dealerships can add to their website to go live. What sets our solution apart is its ability to process and utilize various data formats. We can train Seezar using multiple sources, from car brochures to location data, transforming them into structured information that enhances the bot’s capabilities.
For example, Seezar’s integration with Google Maps allows it to provide real-time distance calculations and directions to dealership locations. When a customer asks about distance, Seezar can instantly calculate it, offer directions, and create a seamless experience. For dealerships with limited digital presence, our team works to build and structure their data by analyzing their online footprint, ensuring Seezar has comprehensive knowledge about their business. This approach has proven particularly valuable for dealerships looking to transition into the AI world. By handling the technical complexities on our end, we’ve made advanced AI technology accessible to businesses at any stage of digital transformation.
Part 4 : Navigating Seezar’s Journey
If you had to build Seezar from scratch again, what would you have done differently?
Building Seezar has been a journey of continuous learning over the past ten months. We quickly realized that creating a powerful AI assistant isn’t just about technology—it’s about understanding real-world challenges in the automotive industry and adapting along the way.
Our step-by-step approach worked well. We started small, learning from every client interaction, and gradually expanded. This helped us understand dealership operations, inventory management, and customer needs in depth. As AI technology evolved from GPT-3.5 to GPT-4, we continuously improved Seezar by integrating new capabilities.
Looking back, we wouldn’t change our approach. Trying to build the perfect AI assistant from day one would have been nearly impossible. Instead, our method of testing, learning, and improving over time was key to making Seezar as advanced and effective as it is today.
How do you envision Seezar evolving to provide even greater value and a more seamless user experience?
Our vision for Seezar is ambitious – we’re building a truly omnichannel AI assistant that accompanies customers throughout their entire automotive journey. From the moment a customer shows interest in a vehicle, whether in the showroom or online, Seezar creates a seamless experience by recognizing them across all touchpoints and maintaining context of their preferences and interactions.
Here’s a small glimpse of Seezar in the future: When you first interact with a dealership, Seezar begins building your profile. If you browse vehicles online after visiting the showroom, Seezar remembers your interests and proactively serves relevant information. It manages the entire customer journey – from scheduling test drives to following up on leads, and ultimately supporting the buying process. Once you become a customer, Seezar transitions into a personal automotive assistant, sending timely service reminders and maintaining ongoing communication.
What makes this particularly powerful for dealerships is Seezar’s ability to manage relationships at scale. For a dealership with thousands of clients, manually tracking service schedules, lease endings, and upgrade opportunities is nearly impossible. Seezar processes this wealth of customer data intelligently, helping dealerships make better decisions while providing personalized attention to each customer. The key is finding the right balance – being helpful without being intrusive, and using AI to transform raw data into meaningful insights that benefit both the dealership and their customers.
What are the most valuable lessons learned from building Seezar that can be applied to future AI initiatives at Seez?
Coming from a design background, my journey into AI development has been both fascinating and enlightening. When I first encountered GPT technology a year ago, it was still a relatively unknown concept. Today, it has revolutionized how we approach problem-solving and innovation in technology.
What makes this journey particularly interesting is how I leveraged AI to learn about AI itself. This self-guided learning approach led to what I playfully call an “AI Inception” – using artificial intelligence to understand and build artificial intelligence solutions. This experience taught me that curiosity is perhaps the most valuable asset in today’s rapidly evolving technological landscape.
The most important lesson I’ve learned is about responsible AI development. While we have access to incredibly powerful tools that “know everything,” the real challenge lies in how we use them. Success in AI development isn’t just about technical capabilities – it’s about building solutions that serve a meaningful purpose and genuinely improve people’s lives. As we stand at the frontier of AI innovation, our focus should remain on creating technology that benefits humanity while being mindful of its ethical implications.
Nader’s final ‘brake’ down:
As I reflect on our journey with Seezar, it’s remarkable to see how we’ve transformed the automotive retail landscape. What started as an ambitious vision to create a human-like AI assistant has evolved into a sophisticated solution that’s reshaping how dealerships interact with their customers.
Seezar has transformed automotive retail through its innovative AI technology, providing instant access to comprehensive vehicle information and reducing recommendation times from 20 minutes to just 7 seconds. The system’s sophisticated ability to handle complex queries, maintain brand integrity, and integrate seamlessly with dealer management systems sets it apart in the industry.
As Seezar evolves into an omnichannel AI assistant, it continues to demonstrate that the future of automotive retail lies in enhancing, rather than replacing, human interaction. By combining artificial intelligence with human expertise, it creates efficient, personalized experiences that benefit both dealerships and customers.
