Neufast, a Singapore-based multilingual recruitment and language training platform, announced on Tuesday, May 6, 2025, that it has raised $1.1 million in a Pre-Series A funding round led by Wavemaker Ventures, the early-stage fund of Wavemaker Partners. Existing backer Wings Capital Ventures also participated in this funding round, which is set to boost product development, market expansion, and enhance Neufast’s proprietary AI capabilities.
Neufast’s early investors include notable firms like SOSV Orbit Startups, Artesian Venture Partners, Her Capital Ventures, AsiaPay Capital, and Connect Beteiligung GmbH based in Vienna, Austria. With the new funding, Neufast plans to expand its engineering team, refine its large language models (LLMs), and build industry-specific integrations, particularly focusing on the insurance and healthcare sectors.
New partnerships with universities and government agencies are also in development, aiming to extend the reach of AI-driven recruitment and training at scale. Agnes Wun, CEng, Co-founder and Chief Executive Officer of Neufast, stated, "Our platform is designed to help organizations hire and upskill talent fairly and efficiently — regardless of language barriers. This new funding enables us to scale our reach and deepen our AI-driven insights, empowering employers to build stronger, more inclusive teams across Asia Pacific.”
Neufast operates as a software as a service (SaaS) company headquartered in Singapore, with operations in Hong Kong and Malaysia, serving clients across the ASEAN region. The firm’s AI-powered multilingual video interview and language training platform supports recruitment in a variety of languages, including Cantonese, Malay, Bahasa Indonesia, Thai, Vietnamese, Tagalog, English, and Mandarin. The platform conducts automated, bias-free job interviews and language assessments in 20 Asian and European languages, significantly reducing recruitment times by up to 88 percent.
Since its launch in May 2020, Neufast has assessed over 40,000 candidates and has helped leading organizations find top talent. Its client roster includes major companies such as Kasikorn Business Technology Group (Thailand), AIA Malaysia, Etiqa Life Insurance Berhad under the Maybank Group in Malaysia, Takaful (Malaysia), Bank Danamon (Indonesia), Carlsberg Asia, the Hong Kong Science and Technology Parks (HKSTP), and Hong Kong’s Urban Renewal Authority.
Beyond recruitment, Neufast’s platform is also utilized for language training. Universities leverage the platform to train students in English oral communication for job interviews, equipping them with critical employability skills. As organizations across the Asia Pacific (APAC) and the European Union (EU) face increasing pressure regarding data governance, cybersecurity, and fair hiring practices, the demand for secure, AI-driven recruitment and language solutions has surged.
Neufast’s ISO 27001:2022 certification and its integration with SAP’s human resource (HR) ecosystem position it uniquely to support regulated industries at scale. Paul Santos, Managing Partner of Wavemaker Partners, commented, “Traditional recruitment processes are slow, costly, and inefficient, causing companies to miss out on top talent. Neufast is changing that with an AI-driven platform that cuts time-to-hire by up to 88% and boosts efficiency by 50 percent.”
Santos further emphasized, “What sets Neufast apart is its ability to localize assessments, incorporating linguistic and cultural nuances to ensure fairer, more accurate evaluations while advancing diversity, equity, and inclusion goals. Beyond automating pre-employment hiring, Neufast is building a next-gen talent management platform for a global, distributed workforce. We’re excited to back their mission to build more connected, skilled teams across Southeast Asia and beyond.”
Meanwhile, five bilingual schools in the Miami area, all part of the LabelFrancEducation network, have received grants from the French Dual Language Fund to support French-English education in U.S. public schools. The funding will cover the cost of DELF (Diplôme d’Études en Langue Française) exams, enhance teaching materials, and support new cultural programming within the schools.
Paxon School for Advanced Studies in Jacksonville also received a grant from the French Heritage Language Program, which supports French-speaking students by funding language classes and cultural activities that preserve their linguistic and cultural roots. Both programs are led by Villa Albertine and made possible through the generous support of the Albertine Foundation.
In the realm of technology, generative AI has made significant strides in recent years, but much of this progress remains concentrated in the English language. Recent research papers have examined the barriers that developers face when trying to close this gap, particularly for languages with less available text data. This lack of large language models (LLMs) for languages with limited online presence threatens to widen existing global divides, potentially cutting off parts of the Global South from transformative technology.
Researchers at the Stanford Institute for Human-Centered AI highlighted that major LLMs tend to underperform for non-English and especially low-resource languages. A survey conducted by Alibaba researchers on over 2,000 multilingual benchmarks found that English remains significantly overrepresented. A new dataset from AI ethics researchers called Shades aims to help developers address cultural context issues by identifying stereotypes and biases across 16 languages.
Low-resource languages, which include Burmese and Swahili, lack both sufficient quantity and quality of digital data necessary for training LLMs. Less than 5% of the approximately 7,000 languages spoken worldwide have meaningful online representation. The Stanford paper evaluated various approaches to addressing this dilemma, including massive multilingual models, regionally specific models, and single-language models.
Ultimately, the researchers recommended strategic investments in R&D for low-resource language AI, promoting global inclusivity in AI research, and ensuring more equitable data ownership. They concluded that "low resourcedness" is not solely a data problem but is also rooted in societal issues, including non-diverse, exclusionary, and exploitative AI research practices.
