Participated · Series A · Jul 2026

Founded in 2022 by Lily Clifford, Brooke Larson, and Ares Geovanos, Rime develops voice AI models tuned for enterprise phone calls. The company operates a recording studio in San Francisco to collect conversational data rather than relying on web-scraped audio and employs a phoneme-based architecture to adapt pronunciations and brand-specific terms. Rime started with separate pipelines for speech-to-text, text-to-speech, and an LLM but is shifting focus toward speech-to-speech models to reduce latency, improve turn-taking, and handle background noise. It serves customers across food service, healthcare, airlines, and fintech and counts Mayo Clinic, Dialpad, Upstart, and Asurion among its enterprise clients. The startup has about 35 employees and recently hired Rafael Valle as Chief Scientist to strengthen its audio understanding and model work. Rime says its training approach helps increase customer call duration, which it credits with winning enterprise contracts.

Participated · Series A · Jun 2026

Coval provides a full‑stack platform for simulating, evaluating, monitoring and improving voice and chat AI agents across the entire agent lifecycle. The company runs tens of millions of evaluations and is trusted by over 60 organizations, including Zoom and Deepgram. Coval says its platform helps enterprises reduce manual QA processes by up to 30x and increase voice agent deployment times by up to 10x. Founded in 2024 by Brooke Hopkins and based in San Francisco, Coval is a Y Combinator graduate. The company has raised a total of $31 million to date, including the $28 million Series A announced in June 2026. With the new funding Coval plans to grow sales and solutions engineering and deepen product capabilities such as simulation, integrations, and human review and monitoring.

Participated · Series C · Jan 2026

Deepgram operates a speech and voice AI platform positioned as infrastructure for developers and enterprises rather than a consumer app. Its platform powers over 200,000 developers and 1,400 organizations, and has processed more than 50,000 years of audio, transcribing over a trillion words. The company has seen rapid APAC growth, with a 96% year‑on‑year increase in API requests and presence in more than 20 APAC markets covering multiple regional languages and dialects. Deepgram plans to open an Asia‑Pacific headquarters in Singapore and expand local sales, support, and low‑latency inference infrastructure to better serve regulated industries and reduce latency. The firm has emphasized low‑latency inference and localized language expertise as key differentiators against large cloud providers and other speech specialists.

Participated · Equity · Dec 2022

Mobile Pathways, founded in 2018 in San Francisco by Jeffrey O'Brien, Poesy Chen, and Bartlomiej Jan Skorupa, builds mobile technology and AI to connect immigrants with reliable legal and employment information. Its core product, Pathfinder, automates immigration status alerts, centralizes vital legal information, and provides data-driven insights for attorneys and advocates. The organization secured $1 million in funding supported by GitLab Foundation, Firedoll Foundation, AlleyCorp, and Fast Forward to advance Pathfinder's AI capabilities. Planned enhancements include leveraging U.S. Citizenship and Immigration Services data to speed work-authorization processes and making case-history reports available to immigration nonprofits for free. AlleyCorp is assisting through its Nonprofit ENG(INE) initiative to help scale Pathfinder’s technical development. Mobile Pathways aims to improve legal outcomes and employment opportunities for underserved immigrant communities by combining justice, technology, and AI. Mobile Pathways operates an online platform that sends automated legal updates to immigrants via text and WhatsApp, built by a team of lawyers, coders, and immigrants. The organization partners with over 140 grassroots immigration nonprofits and reaches nearly 4 million immigrants. It provides digital transformation services including mass text updates and data collection from government agencies, nonprofit partners, and individuals. Mobile Pathways has compiled roughly 100 million rows of immigration data and uses trend analysis to surface systemic issues such as asylum seekers waiting about 720 days for a court date. The group is expanding its use of data and machine learning to articulate immigration court outcomes and increase trust in the immigration process. The newly raised funding will support improvement and expansion of digital transformation services for partner nonprofits.

Participated · Equity · Nov 2022

Frame AI is a next-generation Customer Intelligence system that uses AI to estimate costs and predict revenue outcomes associated with complex customer interactions. The platform consolidates and enriches data across helpdesk, call center, CRM, and other channels to deliver insights for Support, Product, and Revenue leaders. Its Dynamic Cost Attribution product combines budget information with AI analysis of support communication and lifecycle events to identify drivers of operational costs and produce “always-on” lifetime value estimates. The company reports accelerating growth approaching 3x year-over-year driven by that new product offering. Frame AI plans to use the new funding to expand sales, deepen its capabilities in combining natural language and event data, and is hiring both locally and remotely to accelerate growth. Recent senior hires and the addition of an independent board member are intended to scale go-to-market teams and address the needs of multiple enterprise stakeholders. Frame AI builds AI/ML tools to help companies capture and understand customer voice across channels. The product plugs into help desk software, call center tooling, CRMs and other customer communication sources to ingest unstructured interactions. It uses natural language understanding to pull out emerging themes and aggregates insights to account and segment levels so customer experience leaders can prioritize actions. Frame AI integrates with customer data platforms such as Segment, Salesforce Customer 360 and Adobe Real-time CDP, converting unstructured customer voice into moments that can be contributed back to structured systems. CEO and co-founder George Davis describes the product as an early warning system and continuous monitoring tool for customer voice. The company says it plans to continue hiring and building and has raised over $10 million to date, including the new Series A.