Participated · Seed · Jun 2026

QOSMIC develops laser-based optical communications infrastructure for space, including ground stations and satellite terminals. The company has field-validated its full optical communication stack over a 10-kilometre terrestrial link and has secured its first commercial deployment with TakeMe2Space. Founded in 2025 by three founders who returned to India after careers abroad, QOSMIC is based in Bengaluru. Following its $3.33 million seed raise led by Accel and Prosus Ventures, the company plans to deliver operational ground stations and satellite terminals to international customers and expand its engineering team. Accel highlighted the technology as faster, more secure, and far cheaper than existing radio-frequency systems.

Led · Series B · Jun 2026

Equal AI launched an Android app that screens unknown calls and shows users why someone is calling, offering quick-reply options that the AI reads to the caller and allowing custom messages. Since its launch last year the app has grown to more than one million monthly active users and over 300,000 daily active users, and it records calls with transcription history and summaries available in-app. The startup uses a mix of speech recognition, automatic speech recognition (ASR), and speech generation models with its own orchestration layer, and has built support for over 10 languages with code-mixing in mind. Equal began as a data-sharing company for financial services and continues to provide data for financial analysis and KYC verification for employers. The company plans to add the ability to screen calls from known numbers, introduce proactive actions (for example texting delivery details with consent or making outbound appointment calls), build an iOS version, and launch a paid subscription tier. Equal AI was founded in 2022 by Keshav Reddy and faces competition from the call-screening products of Google, Apple, and incumbents like Truecaller.

Participated · Series A · Apr 2026

Deccan AI Experts operates a global network of experts who label data, review model outputs, and run real‑world evaluations to help build accurate and reliable AI systems. The company supports projects across code, agents, and the physical world and works with several leading technology firms, including a majority of the “Mag 7.” Deccan AI reported 10x growth in the prior year and says it has more than 1M contributors across its community. It is Bay Area–based while concentrating much of its workforce in India to manage quality in a fragmented training market. Following its $25M Series A, the company plans to invest heavily in processes, tooling, and people and to accelerate hiring across the U.S. and India. Deccan AI is focused on expanding work in enterprise and physical intelligence and deeper, high‑judgment evaluation projects for frontier models.

Participated · Series A · Mar 2026

Founded in October 2024, Deccan AI supplies post-training services including data generation, expert feedback, evaluations, and building reinforcement learning environments. It offers products such as an evaluation suite called Helix and an operations automation platform, and serves both frontier AI labs and enterprises. The company has onboarded about 10 customers and counts Google DeepMind and Snowflake among its clients, running a couple dozen active projects at a time. Deccan operates from the San Francisco Bay Area with a large operations team in Hyderabad, employs roughly 125 people, and leverages a network of more than 1 million contributors, of whom 5,000 to 10,000 are typically active each month. The startup grew 10x over the past year and is at a double-digit million-dollar revenue run rate, with about 80% of revenue coming from its top five customers. Contributor pay on the platform ranges from about $10 to $700 per hour, with top contributors earning up to $7,000 a month.

Led · Seed · Mar 2026

Founded in 2023, Zapia offers an AI assistant that lets users schedule messages, manage calendars and reminders, transcribe voice notes, search for products or services, and book reservations directly from WhatsApp or its standalone app. The service has quickly scaled to more than 6 million users across Argentina, Brazil, Colombia, Mexico, Peru, and Uruguay. Revenue figures were not disclosed, but the company’s rapid user adoption demonstrates strong market traction in Latin America. Zapia plans to deploy its newly raised capital to launch Zapia Max, an upgraded assistant capable of autonomously executing tasks across websites and applications. By embedding deeper automation and expanding platform reach, the company aims to solidify its position as a leading productivity tool in the region. The seed funding round highlights investor confidence in Zapia’s growth potential and product roadmap.