Tomato.ai

Garry Tan, CEO of Y Combinator, recently shared his perspective on #voiceAI. He said, 'If I were a macro trader, I’d be very bearish on countries that rely heavily on call centers.' Let’s unpack that—and I’ll share my take on what’s next for voice AI. In this 2 minute video, I break down why the ultimate voice experience isn’t just about automation—it’s about seamless collaboration between virtual agents and human agents. Here’s the key: The winning strategy isn’t either/or. It’s a hybrid approach where virtual agents handle routine interactions, and humans step in seamlessly—using the same voice—to close the gap when things get complex. This model unlocks: ? A consistent customer experience ? Greater efficiency for businesses ? A robust solution that balances scale and quality Watch the video and let me know your opinion. #VoiceAI #Automation #FutureOfWork #CCAI #VoiceBots

The Pakistani accent is one of the priority accents we soften at Tomato.ai. Learn about the rich history of the Urdu language and the various regional accents in Pakistan in our latest blog post. #accents #PakistaniAccent #accentAI #AccentNeutralization

Let me help you unpack AI in the call center space There are two key areas where AI is making waves: 1?? Automating voice conversations 2?? Augmenting human agents with AI At Tomato.ai, we focus on augmenting agents, but today, I want to spotlight the automation side—specifically, Salesforce's recent acquisition of Tenyx, a startup specializing in voice virtual agents. Why did Salesforce make this move? Tenyx brought two standout capabilities to the table: Hallucination prevention: They mastered reducing off-track or irrelevant AI responses, ensuring accurate and contextually relevant interactions. Natural conversation handling: Their AI excelled at managing interruptions, seamlessly adapting to customer inputs mid-conversation. These innovations make voice AI not just functional but exceptional. Watch the video and let me know what you think about the future of call center AI! #AI #ContactCenters #VoiceAI #Innovation #CustomerExperience

Have you ever been angry on a customer support call? Most of us have been there. Something’s gone wrong—something’s broken—and we’re triggered. Unfortunately, the person on the other end of the line, the support agent, deals with this daily. The constant wave of emotions, often directed at them, can take a toll. Many leave their roles because it’s just too much to handle day in and day out. But there’s hope. A project out of Japan is developing a fascinating solution: real-time suppression of customer anger during calls. By softening the tone of angry voices, the system helps agents stay calm, leading to faster resolutions and happier outcomes. It’s a win-win. This aligns closely with what we’re building at Tomato.ai. Our accent-softening technology reduces friction by making conversations feel more familiar and comfortable for both sides. The goal? Smoother communication, better understanding, and less frustration—for everyone. What do you think about these emerging innovations? Let’s talk. #AI #CustomerExperience #ContactCenter #Innovation #FutureOfWork #CustomerSupport #AgentWellness #ArtificialIntelligence #TechForGood #CommunicationTools

Key insights from the “Scaling AI Innovation” invite only lunch yesterday, with the GenAI Dir at WhatsApp, the GenAI Product Lead at TikTok, NVIDIA GenAI Startup Sales Exec, etc. - There’s a massive gap between genAI capEX and profits. When will that gap start to close is a big open question - There’s significant underutilization of committed GPUs. If say it’s with training GPUs consider opportunistically using those for inference - There a slew of companies that help with optimizing GPU usage to get more out of lower cost and powered GPUs, which saves money - Conpanies often provision more powerful GPUs than needed for their use case, and overspend I can attest to thst last point, as we tried to go from a100s to h100s and realized soon after that with our size speech AI model, which is much smaller than typical LLMs, we got more bang for our buck from A100s. H100s where about double the cost but only gave us 35% more compute speed. Luckily we were able to back out of our H100 commit, and revert back to using A100s. Thanks IREN for hosting the lunch! Enjoyed chatting with Vivek Jurangal Damien Thioulouse, Nicholas Vitro, Marc Andrew, Dan Lesovodski, Jack Zhang P.S. if anyone wants to hire me to photograph your events let me know ?? #GenAI #ScalingAI #GPUs #AIOptimization #AI

Here are the top three requirements when evaluating an AI accent-softening vendor: naturalness, intelligibility, and latency. When evaluating a vendor, make sure you look at these closely: - Naturalness – Ask yourself: Does the softened voice still sound human? If it’s too robotic, it won’t work across your agent teams. - Intelligibility – The goal is to improve clarity, not lose it. Make sure the softened version maintains a high word recognition rate without altering what’s actually said. - Latency – Every millisecond counts in live conversations. The lower the latency, the smoother the flow. Your agents need an AI that respects real-time interactions without disrupting conversational dynamics. In my view, these are the non-negotiables for any effective accent-softening solution. If you’re exploring this tech, check that all three are up to standard. #AI #AccentSoftening #ContactCenterInnovation #CustomerExperience

Here’s the real problem with an on-device accent softening AI solution: it’s limited by memory and compute power. When agents are using low-powered PCs, they run into two major issues: 1?? Memory limitations: As multiple processes kick off, memory runs low, which affects voice processing quality and consistency. 2?? Reduced AI capabilities: To fit the AI on these devices, we have to miniaturize the model—resulting in a less powerful, less advanced AI. This sacrifices naturalness and clarity in voice quality. The Solution? A cloud-based AI model like Tomato.ai. By moving the AI to the cloud, we get: More processing power without local device constraints A fuller AI model using the latest advancements Low latency, delivering natural, intelligible voice It’s time to focus on solutions that truly empower agents to provide the best experience. #AI #CloudSolutions #CallCenterAI #CustomerExperience #Innovation #onDeviceAI

At Tomato.ai, we’ve removed integration complexity. One of the biggest hurdles I faced back when I was at Google was managing integrations across a diverse set of vendors—Avaya, Genesys, Cisco, you name it. Getting everything to work together often felt like an uphill battle. Our Tomato.ai solution takes just two minutes to set up and integrates effortlessly with your existing soft phone, VoIP, or CRM vendors. Whatever you’re using, it just works. Here’s how it works: Install our solution. No special configuration needed. It sits transparently in front of your existing systems, capturing audio directly from the mic using a secure sound driver. The audio is enhanced and securely transmitted back to your systems—without any storage on our side. If you’re looking for powerful, hassle-free voice enhancement that integrates into your existing setup, Tomato.ai is built for you. To learn more about our solution send me a DM. #AI #ContactCenter #Innovation #VoiceTech #TomatoAI #CustomerExperience #SeamlessIntegration

We are looking to buy accented voice recordings! Please contact us if you have such data or know someone that might have it. #data #aiData #callCenterData

?? Stay Ahead with the Latest in CX AI! ?? The newest CX AI Newsletter is packed with insights to keep you informed on the evolving world of customer experience and generative AI. ?? Gartner’s 2024 Magic Quadrant for CCaaS is here! Gartner ranks the top contact center solutions and explores the balance between human and virtual agents across platforms. ?? AI-Free Contact Center Trends - Not every trend is driven by AI. TechRepublic highlights alternative innovations shaping the contact center landscape. ?? Overcoming 2025’s Challenges - High turnover, omnichannel frustrations, and the pressure for quick AI ROI—Zoom’s CX Ecosystem lead shares practical tips for navigating it all. ?? AI in Action - Learn from six companies leading the way with AI-powered call centers, transforming customer support one interaction at a time. ?? AI Accent Neutralization in the Philippines - This game-changing technology is set to elevate the BPO industry by creating clearer, smoother customer interactions. ?? Read the latest blog: Voice Recognition vs. Accent Translation in Call Centers https://lnkd.in/gQ_GYRn2