Latest news with #Moshi
Yahoo
15-05-2025
- Business
- Yahoo
Energetic Tacoma ramen restaurant abruptly closed. Here's what we know
Sister restaurants Indo Asian Street Eatery and Moshi Ramen Bar, which have served Tacoma's Stadium District since 2015 and 2018, respectively, are in the process of being sold to new owners. Only one will survive the transition, leaving the staff distraught at the loss of their jobs and of a space that had become known as a safe haven for the queer community. Moshi, with various styles of the Japanese noodle soup, izakaya-inspired dishes, fun cocktails and an energetic atmosphere, was one of the only dedicated ramen shops in the area. Its last day of service was Saturday, May 10. On Monday, incoming co-owner Kevin Merrell informed the approximately 16 employees that Moshi was permanently closed. He plans to renovate and replace the concept with 'authentic Thai,' he confirmed to The News Tribune on Tuesday. He and his wife Thip finalized a deal this week to buy the restaurant from Yu Nanakornphanom and Buoy Ngov. The Merrells are also buying neighboring Indo, but that sale is separate, both parties said. Merrell declined to reveal the name of the new Thai restaurant or further menu details beyond that the price point would be similar to Indo's. Despite the potential conceptual overlap and physical proximity — the two restaurants share a basement for prep and storage — he added only that it will 'be different enough that it makes sense.' Next door, he said, 'Indo will continue to be Southeast fusion, just as it is.' Merrell was raised in Tacoma and has been a business owner for decades, he said, including a local landscaping company and a motorcycle tour company in Thailand, where he lived for eight years and met Thip. After moving back to the Pacific Northwest, he said, she wanted to 'get back in the restaurant business.' They filed for business licenses in January and applied for updated liquor licenses in February, according to state records. The Indo-Moshi sale has been in the works since last year. Nanakornphanom and Ngov had stepped away from day-to-day operations to spend more time with family, caring for their own health and their aging parents. They left both restaurants in the hands of managers, but the couple said they didn't share details of the sale with most employees due to confidentiality terms of the purchase agreement. That void left several former and current employees disappointed in the situation that transpired over the last eight months, according to several who spoke with The News Tribune. Thip Merrell had been working in the kitchen at Indo, they said, and Kevin Merrell began formally managing Indo in December. Moshi's employees were by then aware of the pending sale but said they were told little else about their fate. 'Officially we knew absolutely nothing. Especially in retrospect, there were some signs,' said Samuel Kirbawy, who had managed the Japanese restaurant since it opened in 2018 with late chef Aaron Grissom at the helm. He felt they had been 'misled into thinking that the restaurant would continue' under its new ownership and wished 'it had been handled in a more respectful way.' Lane Parrish worked at the Japanese restaurant for four years and loved it, they said in a phone call on Tuesday. The pay was fair with consistent raises for tenure, but beyond that, 'The biggest thing was really just the environment,' they said. 'Everyone was welcome. We were able to create a welcoming environment because we were able to be our genuine selves.' 'We all would've liked to have said goodbye to our community,' added Kirbawy, who started a GoFundMe to support the cooks, bartenders, servers and hosts who were paid hourly. 'Moshi was providing employment for a lot of marginalized people. It's a pretty big blow to some people.' Some local business owners have encouraged Moshi's staff to consider applying for positions at their restaurants, they said. A few also chimed into Instagram posts to extend that invitation. In phone calls this week, Nankornphanom and Ngov said that Moshi had not been financially sustainable for some time. They were struggling to make payroll and had considered closing it earlier, they said, but the sale process complicated that picture. 'It wasn't personal with anything. We just couldn't make it anymore,' said Nankornphanom. In some ways, they continued, Indo and Moshi had ballooned into much bigger and more complicated enterprises than the mom-and-pop, 'real family restaurant' idea that started their entrepreneurial journey in 2015. 'We're sad, too,' said Ngov. In the future they might consider a small, family-run restaurant but for now will focus on family. At the Monday meeting, Merrell offered employees the opportunity to apply for a position at the new Thai restaurant, which he anticipates introducing this summer. He confirmed they would need to pursue unemployment insurance in the meantime. 'We made that offer. No one took it,' he told The News Tribune the next day, noting the common practice of what's known in employment law as 'technical termination' during the sale of a business. Existing employees must be terminated if they are to be rehired by the buyer at similar wages and in similar jobs. To avoid being categorized as a loss of employment, the rehiring must happen within six months of the sale, according to the U.S. Department of Labor. 'We'll be down for a little bit rebranding the location,' added Merrell. ▪ 110 N. Tacoma Ave., Tacoma, 253-503-3527, ▪ Wednesday-Saturday 11 a.m.-9 p.m., Sunday 11 a.m.-8 p.m. ▪ New owners will maintain Indo Asian; new Thai restaurant coming to Moshi space next door
IOL News
07-05-2025
- Business
- IOL News
Public servants' housing allowance increased: new rates effective from last month
The Kanku Road Housing Development in Isipingo. Image: Isipingo ratepayers The Department of Public Service and Administration(DPSA) has officially announced an upward adjustment to the housing allowance for public servants. According to Circular No. 15 of 2025, the revised allowance, implemented per the Public Service Coordinating Bargaining Council (PSCBC) Resolution 1 of 2025, took effect at the beginning of last month. Effective from that date, the monthly housing allowance for qualifying public servants will increase from R1 784.55 to R1 900.00. The housing allowance, as contained in the Public Service Coordinating Bargaining Council (PSCBC) Resolution 7 of 2015 (clause 4.6), provides that the amount of the housing allowance shall be adjusted annually based on the average Consumer Price Index (CPI) for the preceding financial year. The policy on housing allowance also outlines specific conditions for employees who do not own a home, which include that tenants with legal rental agreements appointed before May 27, 2015, will continue to receive a direct R900 monthly allowance. The remaining R1000 will be saved in the Government Employees Housing Scheme's (GEHS) Individual-Linked Savings Facility (ILSF), which is managed by the National Treasury. For employees appointed on or after May 27, 2015, will have the full R1 900 housing allowance diverted into the ILSF, promoting savings towards future homeownership. Earlier this year, Moses Moshi, the spokesperson for the DPSA, told this publication that every month, an average of R164 million of ILSF savings (Housing Allowance benefit) is processed, albeit not exclusively for homeownership. He said this included employees who withdraw because of retirement, being medically boarded, end of contract, death or reversals. Before 2015, when the ILSF was not yet in existence, the housing allowance benefit was being paid directly to employees irrespective of whether they are homeowners or tenants. From March 2016 to the end of January this year, when employees who are not homeowners had their housing allowance benefit saved in the ILSF, the cumulative savings amounted to over R28 billion as at the end of January this year. From 2015 to the end of January this year 555 892 housing allowance withdrawals were processed through the ILSF to the value of R11 billion. The secular said that PERSAL has been directed to effect the adjustments on the system to ensure seamless implementation, while national and provincial departments are expected to fund these increases from their existing budget allocations. The Government Employees Housing Scheme (GEHS), along with the annual subsidy adjustments, reflects the government's ongoing commitment to improving the living conditions of public servants and to encouraging responsible homeownership through structured savings mechanisms. According to Moshi, before 2015, public servants received the housing allowance as part of the condition of service benefit for years. However, he said this allowance was not being used by several employees to acquire homes. Hence, in 2015, the government (as the employer) and labour signed the Public Service Coordinating Bargaining Council (PSCBC) Resolution 7 of 2015, to establish a (GEHS). According to the department, the challenges concerning the housing allowance benefit, which impact on homeownership by government employees, was that the majority of employees who receive the housing allowance benefit are those at levels 1 – 7 (as at 31 December 2024 equalling to 638 836) and mostly did not qualify for mortgage bonds at commercial banks and also, do not qualify for social housing programmes of government. 'The current finance products are too expensive for most government employees who fall within the gap market. As such, a market product needs to be developed for this segment of employees so that they can also become homeowners. "This would require collaboration between government, government entities, development finance institutions and the commercial housing finance sector to develop such products that would accommodate these employees market.' Video Player is loading. Play Video Play Unmute Current Time 0:00 / Duration -:- Loaded : 0% Stream Type LIVE Seek to live, currently behind live LIVE Remaining Time - 0:00 This is a modal window. Beginning of dialog window. Escape will cancel and close the window. Text Color White Black Red Green Blue Yellow Magenta Cyan Transparency Opaque Semi-Transparent Background Color Black White Red Green Blue Yellow Magenta Cyan Transparency Opaque Semi-Transparent Transparent Window Color Black White Red Green Blue Yellow Magenta Cyan Transparency Transparent Semi-Transparent Opaque Font Size 50% 75% 100% 125% 150% 175% 200% 300% 400% Text Edge Style None Raised Depressed Uniform Dropshadow Font Family Proportional Sans-Serif Monospace Sans-Serif Proportional Serif Monospace Serif Casual Script Small Caps Reset restore all settings to the default values Done Close Modal Dialog End of dialog window. Advertisement Next Stay Close ✕ Ad Loading
Yahoo
18-02-2025
- Business
- Yahoo
Deepgram Achieves Key Milestone on Path to Delivering Next-Gen, Enterprise-Grade Speech-to-Speech Architecture
Pioneering Achievement Delivers Speech-To-Speech Technology Without Intermediate Text Representations, Setting the Stage for Fully Fluid, Human-Like Enterprise Voice AI Applications SAN FRANCISCO, February 18, 2025--(BUSINESS WIRE)--Deepgram, the leader in enterprise-grade speech AI, today announced a significant technical achievement in speech-to-speech (STS) technology for enterprise use cases. The company has successfully developed a speech-to-speech model that operates without relying on text conversion at any stage, marking a pivotal step toward the development of contextualized end-to-end speech AI systems. This milestone will enable fully natural and responsive voice interactions that preserve nuances, intonation, and emotional tone throughout real-time communication. When fully operationalized, this architecture will be delivered to customers via a simple upgrade from our existing industry-leading architecture. By adopting this technology alongside Deepgram's full-featured voice AI platform, companies will gain a strategic advantage, positioning themselves to deliver cutting-edge, scalable voice AI solutions that evolve with the market and outpace competitors. Advancements Over Existing Architectures Existing speech-to-speech (STS) systems are based on architectures that process speech through sequential stages, such as speech-to-text, text-to-text, and text-to-speech. These architectures have become the standard for production deployments for their modularity and maturity, but eliminating text as an intermediary offers opportunities to improve latency and better preserve emotional and contextual nuances. Meanwhile, multimodal LLMs like Gemini, GPT-4o, and Llama have evolved beyond text-only capabilities to accept additional inputs such as images, videos, and audio. However, despite these advancements, they struggle to capture the fluidity and nuance of human-like conversation. These models still rely on a turn-based framework, where audio input is tokenized and processed within a textual domain, restricting real-time interactivity and expressiveness. To advance the frontier of speech AI, Deepgram is setting the stage for end-to-end STS models, which offer a more direct approach by converting speech to speech without relying on text. Recent research on speech-to-speech models, such as Hertz and Moshi, has highlighted the significant challenges in developing models that are robust and reliable enough for enterprise use cases. These difficulties stem from the inherent complexities of modeling conversational speech and the substantial computational resources required. Overcoming these hurdles demands innovations in data collection, model architecture, and training methodologies. Delivering Speech-to-Speech with Latent Space Embeddings Deepgram is transforming speech-to-speech modeling with a new architecture that fuses the latent spaces of specialized components, eliminating the need for text conversion between them. By embedding speech directly into a latent space, Deepgram ensures that important characteristics such as intonation, pacing, and situational and emotional context are preserved throughout the entire processing pipeline. What sets Deepgram apart is its approach to fusing the hidden states—the internal representations that capture meaning, context, and structure—of each individual function: Speech-to-Text (STT), Large Language Model (LLM), and Text-to-Speech (TTS). This fusion is the first step toward training a controllable single, true end-to-end speech model, enabling seamless processing while retaining the strengths of each best-in-class component. This breakthrough has significant implications for enterprise applications, facilitating more natural conversations while maintaining the control and reliability businesses require. "This achievement represents a fundamental shift in how AI systems can process and respond to human speech," said Scott Stephenson, CEO and Co-founder of Deepgram. "By eliminating text as an intermediate step, we're preserving crucial elements of communication and maintaining the precise control that enterprises need for mission-critical applications." This technical advancement builds on Deepgram's expertise in enterprise speech AI, with over 200,000 developers using its platform, more than 50,000 years of audio processed, and over 1 trillion words transcribed. Key benefits of the new architecture include: Optimized latency design for faster, more responsive interactions Enhanced naturalness, preserving emotional context and conversational nuances Native ability to handle complex, multi-turn conversations Unified, end-to-end training across the entire model, creating a more cohesive and inherently adaptive system that fine-tunes its understanding and response generation directly in the audio space. Utilizing Transfer Learning for Cost-Efficient, High-Accuracy Speech-to-Speech Deepgram's research in the space is accelerated by its use of transfer learning and best-in-class pre-trained models, allowing it to achieve high accuracy with significantly less training data than traditional methods. Without latent techniques, training a model at the scale needed for speech-to-speech would require over 80 billion hours of audio—more than humanity has ever recorded. However, Deepgram's latent space embeddings and transfer learning approach achieve superior comprehension while significantly reducing costs, maintaining interpretability, and accelerating enterprise deployment. This efficiency enables Deepgram to deliver scalable, end-to-end speech AI that meets the demands of real-world voice applications. Empowering Developers with Full Debuggability One of the requirements in enterprise speech-to-speech modeling is the ability to understand and troubleshoot each step of the process. This is particularly challenging when text conversion between steps isn't involved, as verifying both the accuracy of the initial perception and the alignment of the spoken output with the intended response is not straightforward. Deepgram recognized this need and addressed it by designing a new architecture that enables debuggability throughout the entire process. This architecture allows developers to inspect and understand how the system processes spoken dialogue. The design incorporates speech modeling of perception, natural language understanding/generation, and speech production, preserving distinct capabilities during training. Through the ability to decode intermediate representations back to text at specific points, developers can gain insight into what the model perceives, thinks, and generates, ensuring its internal representation aligns with the model output and stays true to the intent of the business user, addressing hallucination concern in scaled business use cases. This capability allows the user to peer into each step throughout generation, helping refine models, improve performance, and deliver more accurate, lifelike, and reliable speech-to-speech solutions. Beyond Speech-to-Speech – A Complete, Enterprise-Ready Voice AI Stack While building an advanced speech-to-speech (STS) model is a major technical achievement, enterprises need more than just a model—they need a complete, scalable platform that ensures seamless deployment, adaptability, and cost efficiency. Deepgram delivers not just cutting-edge STS technology, but an enterprise-ready infrastructure designed for real-world applications. Seamless Integration & Continuous Improvement – Once Deepgram's end-to-end STS model moves to production, businesses will be able to adopt this breakthrough directly through our developer-friendly voice agent API from within the current Deepgram platform. Through continued innovation, enterprises will benefit from the latest advancements, ensuring seamless integration and a future-proof platform for their voice AI applications. Enterprise-Grade Performance & Cost Efficiency – Built for low customer COGS, our platform enables enterprises to deploy high-performance voice AI without excessive costs. This ensures scalability, whether for customer service automation, real-time voice agents, or multilingual applications. Full-Featured Platform and High-Performance Runtime – Deepgram's platform includes powerful capabilities such as: Adaptability - Dynamically fine-tune models for specific industry language, ensuring high accuracy across diverse applications without needing constant retraining. Automation - Streamline transcription, model updates, and data processing, reducing overhead and accelerating deployment. Synthetic data generation - Generate synthetic voice data to improve model training, even with limited real-world data, enhancing accuracy for niche use cases. Data curation - Clean, manage, and organize training data to ensure high-quality, relevant input, improving model performance. Model hot-swapping - Seamlessly switch between different models to optimize performance for specific tasks. Integrations - Effortlessly integrate Deepgram's voice AI with cloud platforms, enterprise systems, and third-party applications, embedding it within existing workflows. With Deepgram, enterprises don't just get speech-to-speech—they get the most advanced, enterprise-ready voice AI platform, designed for real-world deployment and long-term innovation. For more information about Deepgram's novel approach for speech-to-speech, read the technical brief. To learn more about Deepgram's suite of voice AI infrastructure, visit Additional Resources: Explore the technical brief on Deepgram's novel speech-to-speech architecture Watch a fun demo of Deepgram's voice agent API Try Deepgram's interactive demo Get $200 in free credits and try Deepgram for yourself About Deepgram Deepgram is the leading voice AI platform for enterprise use cases, offering speech-to-text (STT), text-to-speech (TTS), and full speech-to-speech (STS) capabilities. 200,000+ developers build with Deepgram's voice-native foundational models – accessed through cloud APIs or as self-hosted / on-premises APIs – due to our unmatched accuracy, low latency, and pricing. Customers include technology ISVs building voice products or platforms, co-sell partners working with large enterprises, and enterprises solving internal use cases. Having processed over 50,000 years of audio and transcribed over 1 trillion words, there is no organization in the world that understands voice better than Deepgram. To learn more, visit read our developer docs, or follow @DeepgramAI on X and LinkedIn. View source version on Contacts PR Contact: Nicole GormanGorman Communications, for DeepgramM: Sign in to access your portfolio
