COMPANY INFO
Pika is a generative AI platform for creating and editing videos, images and audio. It provides its own models alongside third-party models in a creative suite and offers tools through web, mobile and API services.
This page is built with Exa, a web search API for AI agents. It is updated every 24 hours.
Pika is a generative AI platform for creating and editing videos, images and audio. It provides its own models alongside third-party models in a creative suite and offers tools through web, mobile and API services.
Official website: https://pika.art
Leads Pika and its vision for accessible AI-powered creation.
Studied mathematics and computer science at Harvard, conducted AI research at Stanford, and worked at Meta, Google, Microsoft and Epic Games.
Leads scientific work on image and video generation.
Conducted generative-model research at Stanford and co-authored the DDIM diffusion-model paper.
Oversees operations at Pika.
Previously a principal at Makers Fund and an engagement manager at McKinsey.
Leads creative direction at Pika.
Previously directed television commercials for approximately ten years.
Leads engineering at Pika.
Previously worked on distributed systems at Google and engineering at Wattpad and Amazon.
Leads Pika AI Self Agent product, engineering and agent core.
Previously worked in core product management at TikTok and Douyin.
This page is built with Exa, a web search API for AI agents. It is updated every 24 hours.
Exa searches the web, extracts relevant source content, and returns structured outputs with citations.
const run = await exa.agent.runs.create({
query: "Research Pika's leadership team, key executives, and founders. Return structured JSON with field-level citations.",
effort: "low",
outputSchema: {
// company, executives, founders
/* ... */
}
});
const result = await exa.agent.runs.pollUntilFinished(run.id);{
"status": "completed",
"output": {
"structured": {
"company": "Pika",
"executives": [
"..."
],
"founders": [
"..."
]
},
"grounding": [
{
"field": "executives[0].title",
"citations": [
{
"url": "..."
}
]
}
]
}
}