AI Dev 25 X NYC | Kay Zhu: How Genspark Built a Super Agent That Scales

Why It Matters

GenSpark’s agentic AI platform dramatically cuts the time required for data analysis, content creation, and presentation building, unlocking productivity gains that could redefine work schedules for knowledge workers and accelerate AI adoption across enterprise functions.

Summary

Kay Zhu, CTO and co‑founder of GenSpark, opened the AI Dev 25 × NYC session by positioning GenSpark as an all‑in‑one, agentic AI workspace aimed at turning white‑collar work into a "three‑day work week" for over a billion knowledge workers. The company, founded two years ago, has raised $160 million across two funding rounds, hit $50 million ARR within five months of launching its Super Agent suite in April 2025, and now serves more than 10 million users globally. Zhu highlighted a rapid product cadence—new features almost every week—including AI‑driven slides, spreadsheets, documents, browsers, drives, podcasts, and the latest AI employee tools such as AI developer, designer, and video editor agents.

The presentation pivoted to live demonstrations with two power users. Real‑estate private‑equity professional John showcased how GenSpark ingested raw Excel rent‑roll data, auto‑generated Python‑based analyses, and produced investor‑ready presentations in minutes—tasks that previously consumed hours. The platform’s "mixture of agents" and "auto‑prompt" capabilities allowed John to generate multiple image variants and detailed reports with minimal prompting, illustrating the system’s ability to orchestrate sub‑agents for data extraction, code execution, and slide creation.

Educator‑entrepreneur Dave illustrated a different use case: after recording a 90‑minute podcast, he fed the transcript into GenSpark, which automatically summarized the content into a four‑minute, dual‑voice audio snippet and generated supporting assets for multiple audience segments. This workflow underscores GenSpark’s promise to repurpose long‑form content quickly, reducing manual editing and transcription effort for creators across industries.

Collectively, the demos emphasized GenSpark’s core value proposition—leveraging autonomous AI agents to automate repetitive, data‑intensive tasks, thereby compressing weeks of work into minutes. The company’s rapid user growth, sizable ARR, and integration into the OpenAI trillion‑token club suggest it is scaling both its technology stack and market adoption, positioning itself as a pivotal infrastructure layer for the next wave of AI‑augmented productivity tools.

Original Description

Kay Zhu, CTO of Mainfunc (Genspark), shared at AI Dev 25 x NYC how Genspark advocates for giving AI agents autonomy to plan rather than rigid workflows, while equipping them with 80+ specialized tools.

His belief: workflows fail on edge cases and accumulate errors, while agents can observe, backtrack, and recover from unexpected situations.

--------------

Join us at AI Dev 26 x San Francisco! Tickets: https://ai-dev.deeplearning.ai/