Building an enterprise AI company on a “foundation of shifting sand” is the central challenge for founders today, according to the leadership at Palona AI. Today, the Palo Alto-based startup\u2014led by former Google and Meta engineering veterans\u2014is making a decisive vertical push into the restaurant and hospitality space with today’s launch of Palona Vision and Palona Workflow. The new offerings transform the company\u2019s multimodal agent suite into a real-time operating system for restaurant operations \u2014 spanning cameras, calls, conversations, and coordinated task execution.The news marks a strategic pivot from the company\u2019s debut in early 2025, when it first emerged with $10 million in seed funding to build emotionally intelligent sales agents for broad direct-to-consumer enterprises. Now, by narrowing its focus to a “multimodal native” approach for restaurants, Palona is providing a blueprint for AI builders on how to move beyond “thin wrappers” to build deep systems that solve high-stakes physical world problems.\u201cYou\u2019re building a company on top of a foundation that is sand\u2014not quicksand, but shifting sand,\u201d said co-founder and CTO Tim Howes, referring to the instability of today\u2019s LLM ecosystem. \u201cSo we built an orchestration layer that lets us swap models on performance, fluency, and cost.\u201dVentureBeat spoke with Howes and co-founder and CEO Maria Zhang in person recently at \u2014 where else? \u2014 a restaurant in NYC about the technical challenges and hard lessons learned from their launch, growth, and pivot.The New Offering: Vision and Workflow as a \u2018Digital GM\u2019For the end user\u2014the restaurant owner or operator\u2014Palona\u2019s latest release is designed to function as an automated “best operations manager” that never sleeps.Palona Vision uses in-store security cameras to analyze operational signals \u2014 such as queue lengths, table turnover, prep bottlenecks, and cleanliness \u2014 without requiring any new hardware.It monitors front-of-house metrics like queue lengths, table turns, and cleanliness, while simultaneously identifying back-of-house issues like prep slowdowns or station setup errors.Palona Workflow complements this by automating multi-step operational processes. This includes managing catering orders, opening and closing checklists, and food prep fulfillment. By correlating video signals from Vision with Point-of-Sale (POS) data and staffing levels, Workflow ensures consistent execution across multiple locations. \u201cPalona Vision is like giving every location a digital GM,\u201d said Shaz Khan, founder of Tono Pizzeria + Cheesesteaks, in a press release provided to VentureBeat. \u201cIt flags issues before they escalate and saves me hours every week.\u201dGoing Vertical: Lessons in Domain ExpertisePalona\u2019s journey began with a star-studded roster. CEO Zhang previously served as VP of Engineering at Google and CTO of Tinder, while Co-founder Howes is the co-inventor of LDAP and a former Netscape CTO. Despite this pedigree, the team\u2019s first year was a lesson in the necessity of focus.Initially, Palona served fashion and electronics brands, creating “wizard” and “surfer dude” personalities to handle sales. However, the team quickly realized that the restaurant industry presented a unique, trillion-dollar opportunity that was “surprisingly recession-proof” but “gobsmacked” by operational inefficiency.”Advice to startup founders: don’t go multi-industry,” Zhang warned. By verticalizing, Palona moved from being a “thin” chat layer to building a “multi-sensory information pipeline” that processes vision, voice, and text in tandem.That clarity of focus opened access to proprietary training data (like prep playbooks and call transcripts) while avoiding generic data scraping. 1. Building on \u2018Shifting Sand\u2019To accommodate the reality of enterprise AI deployments in 2025 \u2014 with new, improved models coming out on a nearly weekly basis \u2014 Palona developed a patent-pending orchestration layer.Rather than being “bundled” with a single provider like OpenAI or Google, Palona\u2019s architecture allows them to swap models on a dime based on performance and cost. They use a mix of proprietary and open-source models, including Gemini for computer vision benchmarks and specific language models for Spanish or Chinese fluency. For builders, the message is clear: Never let your product’s core value be a single-vendor dependency.2. From Words to \u2018World Models\u2019The launch of Palona Vision represents a shift from understanding words to understanding the physical reality of a kitchen. While many developers struggle to stitch separate APIs together, Palona\u2019s new vision model transforms existing in-store cameras into operational assistants.The system identifies “cause and effect” in real-time\u2014recognizing if a pizza is undercooked by its “pale beige” color or alerting a manager if a display case is empty. “In words, physics don’t matter,” Zhang explained. “But in reality, I drop the phone, it always goes down… we want to really figure out what’s going on in this world of restaurants”.3. The \u2018Muffin\u2019 Solution: Custom Memory ArchitectureOne of the most significant technical hurdles Palona faced was memory management. In a restaurant context, memory is the difference between a frustrating interaction and a “magical” one where the agent remembers a diner\u2019s “usual” order.The team initially utilized an unspecified open-source tool, but found it produced errors 30% of the time. “I think advisory developers always turn off memory [on consumer AI products], because that will guarantee to mess everything up,” Zhang cautioned.To solve this, Palona built Muffin, a proprietary memory management system named as a nod to web “cookies”. Unlike standard vector-based approaches that struggle with structured data, Muffin is architected to handle four distinct layers:Structured Data: Stable facts like delivery addresses or allergy information.Slow-changing Dimensions: Loyalty preferences and favorite items.Transient and Seasonal Memories: Adapting to shifts like preferring cold drinks in July versus hot cocoa in winter.Regional Context: Defaults like time zones or language preferences.The lesson for builders: If the best available tool isn’t good enough for your specific vertical, you must be willing to build your own.4. Reliability through \u2018GRACE\u2019In a kitchen, an AI error isn’t just a typo; it\u2019s a wasted order or a safety risk. A recent incident at Stefanina\u2019s Pizzeria in Missouri, where an AI hallucinated fake deals during a dinner rush, highlights how quickly brand trust can evaporate when safeguards are absent.To prevent such chaos, Palona\u2019s engineers follow its internal GRACE framework:Guardrails: Hard limits on agent behavior to prevent unapproved promotions.Red Teaming: Proactive attempts to “break” the AI and identify potential hallucination triggers.App Sec: Lock down APIs and third-party integrations with TLS, tokenization, and attack prevention systems.Compliance: Grounding every response in verified, vetted menu data to ensure accuracy.Escalation: Routing complex interactions to a human manager before a guest receives misinformation.This reliability is verified through massive simulation. “We simulated a million ways to order pizza,” Zhang said, using one AI to act as a customer and another to take the order, measuring accuracy to eliminate hallucinations.The Bottom LineWith the launch of Vision and Workflow, Palona is betting that the future of enterprise AI isn’t in broad assistants, but in specialized “operating systems” that can see, hear, and think within a specific domain. In contrast to general-purpose AI agents, Palona\u2019s system is designed to execute restaurant workflows, not just respond to queries \u2014 it’s capable of remembering customers, hearing them order their “usual,” and monitoring the restaurant operations to ensure they deliver that customer the food according to their internal processes and guidelines, flagging whenever something goes wrong or crucially, is about to go wrong.For Zhang, the goal is to let human operators focus on their craft: “If you’ve got that delicious food nailed… we\u2019ll tell you what to do.”<\/p>\n","protected":false},"excerpt":{"rendered":"
Building an enterprise AI company on a “foundation of shifting sand” is the central challenge for founders today, according to the leadership at Palona AI<\/a>. <\/p>\n Today, the Palo Alto-based startup\u2014led by former Google and Meta engineering veterans\u2014is making a decisive vertical push into the restaurant and hospitality space with today’s launch of Palona Vision and Palona Workflow. <\/p>\n The new offerings transform the company\u2019s multimodal agent suite into a real-time operating system for restaurant operations \u2014 spanning cameras, calls, conversations, and coordinated task execution.<\/p>\n The news marks a strategic pivot from the company\u2019s debut in early 2025<\/a>, when it first emerged with $10 million in seed funding<\/a> to build emotionally intelligent sales agents for broad direct-to-consumer enterprises. <\/p>\n Now, by narrowing its focus to a “multimodal native” approach for restaurants, Palona is providing a blueprint for AI builders on how to move beyond “thin wrappers” to build deep systems that solve high-stakes physical world problems.<\/p>\n \u201cYou\u2019re building a company on top of a foundation that is sand\u2014not quicksand, but shifting sand,\u201d said co-founder and CTO Tim Howes, referring to the instability of today\u2019s LLM ecosystem. \u201cSo we built an orchestration layer that lets us swap models on performance, fluency, and cost.\u201d<\/p>\n VentureBeat spoke with Howes and co-founder and CEO Maria Zhang in person recently at \u2014 where else? \u2014 a restaurant in NYC about the technical challenges and hard lessons learned from their launch, growth, and pivot.<\/p>\n For the end user\u2014the restaurant owner or operator\u2014Palona\u2019s latest release is designed to function as an automated “best operations manager” that never sleeps.<\/p>\n Palona Vision uses in-store security cameras to analyze operational signals \u2014 such as queue lengths, table turnover, prep bottlenecks, and cleanliness \u2014 without requiring any new hardware.<\/p>\n It monitors front-of-house metrics like queue lengths, table turns, and cleanliness, while simultaneously identifying back-of-house issues like prep slowdowns or station setup errors.<\/p>\n Palona Workflow complements this by automating multi-step operational processes. This includes managing catering orders, opening and closing checklists, and food prep fulfillment. By correlating video signals from Vision with Point-of-Sale (POS) data and staffing levels, Workflow ensures consistent execution across multiple locations. <\/p>\n \u201cPalona Vision is like giving every location a digital GM,\u201d said Shaz Khan, founder of Tono Pizzeria + Cheesesteaks, in a press release provided to VentureBeat. \u201cIt flags issues before they escalate and saves me hours every week.\u201d<\/p>\n Palona\u2019s journey began with a star-studded roster. CEO Zhang previously served as VP of Engineering at Google and CTO of Tinder, while Co-founder Howes is the co-inventor of LDAP and a former Netscape CTO. <\/p>\n Despite this pedigree, the team\u2019s first year was a lesson in the necessity of focus.<\/p>\n Initially, Palona served fashion and electronics brands, creating “wizard” and “surfer dude” personalities to handle sales. However, the team quickly realized that the restaurant industry presented a unique, trillion-dollar opportunity that was “surprisingly recession-proof” but “gobsmacked” by operational inefficiency.<\/p>\n “Advice to startup founders: don’t go multi-industry,” Zhang warned. <\/p>\n By verticalizing, Palona moved from being a “thin” chat layer to building a “multi-sensory information pipeline” that processes vision, voice, and text in tandem.<\/p>\n That clarity of focus opened access to proprietary training data (like prep playbooks and call transcripts) while avoiding generic data scraping. <\/p>\n 1. Building on \u2018Shifting Sand\u2019<\/b><\/p>\n To accommodate the reality of enterprise AI deployments in 2025 \u2014 with new, improved models coming out on a nearly weekly basis \u2014 Palona developed a patent-pending orchestration layer.<\/p>\n Rather than being “bundled” with a single provider like OpenAI or Google, Palona\u2019s architecture allows them to swap models on a dime based on performance and cost. <\/p>\n They use a mix of proprietary and open-source models, including Gemini for computer vision benchmarks and specific language models for Spanish or Chinese fluency. <\/p>\n For builders, the message is clear: Never let your product’s core value be a single-vendor dependency.<\/p>\n 2. From Words to \u2018World Models\u2019<\/b><\/p>\n The launch of Palona Vision represents a shift from understanding words to understanding the physical reality of a kitchen. <\/p>\n While many developers struggle to stitch separate APIs together, Palona\u2019s new vision model transforms existing in-store cameras into operational assistants.<\/p>\n The system identifies “cause and effect” in real-time\u2014recognizing if a pizza is undercooked by its “pale beige” color or alerting a manager if a display case is empty. <\/p>\n “In words, physics don’t matter,” Zhang explained. “But in reality, I drop the phone, it always goes down… we want to really figure out what’s going on in this world of restaurants”.<\/p>\n 3. The \u2018Muffin\u2019 Solution: Custom Memory Architecture<\/b><\/p>\n One of the most significant technical hurdles Palona faced was memory management. In a restaurant context, memory is the difference between a frustrating interaction and a “magical” one where the agent remembers a diner\u2019s “usual” order.<\/p>\n The team initially utilized an unspecified open-source tool, but found it produced errors 30% of the time. “I think advisory developers always turn off memory [on consumer AI products], because that will guarantee to mess everything up,” Zhang cautioned.<\/p>\n To solve this, Palona built Muffin, a proprietary memory management system named as a nod to web “cookies”. Unlike standard vector-based approaches that struggle with structured data, Muffin is architected to handle four distinct layers:<\/p>\n Structured Data: Stable facts like delivery addresses or allergy information.<\/p>\n<\/li>\n Slow-changing Dimensions: Loyalty preferences and favorite items.<\/p>\n<\/li>\n Transient and Seasonal Memories: Adapting to shifts like preferring cold drinks in July versus hot cocoa in winter.<\/p>\n<\/li>\n Regional Context: Defaults like time zones or language preferences.<\/p>\n<\/li>\n<\/ul>\n The lesson for builders: If the best available tool isn’t good enough for your specific vertical, you must be willing to build your own.<\/p>\n 4. Reliability through \u2018GRACE\u2019<\/b><\/p>\n In a kitchen, an AI error isn’t just a typo; it\u2019s a wasted order or a safety risk. A recent incident at Stefanina\u2019s Pizzeria in Missouri, where an AI hallucinated fake deals during a dinner rush<\/a>, highlights how quickly brand trust can evaporate when safeguards are absent.<\/p>\nThe New Offering: Vision and Workflow as a \u2018Digital GM\u2019<\/b><\/h3>\n
Going Vertical: Lessons in Domain Expertise<\/b><\/h3>\n
\n