HomeCompaniesKestral
Kestral

We automate your entire product development lifecycle.

Kestral makes it clear what your customers want most, prioritizes those tasks for highest leverage, and closes the loop with those customers. We (Brian and Bernard) have been close friends since university, and most recently worked together at Asana, leading AI and Authorization efforts. There we realized AI can completely disrupt the $50B Work Management Market, but it required a fresh start. What work we even choose to manage in the first place is critical, and the most important insights from customers are silo-ed in sales/CS while R&D gives their best shot at roadmaps meant to deliver impact. Significant requests sit in a backlog, duplicate asks flood the system, months wasted on "strategic investments". We want to bridge this gap. This message is resonating with heads of sales, heads of product, CEOs, as well as individuals, all trying to make sure the most important things get done, with less wasted effort and more revenue/retention. Kestral works by automatically take in customer data like sales call transcripts & feature request messages and a team of agents: 1. analyze them for actionable takeaways, deal blockers, and nice to haves 2. consolidate them and manage all the asks with citations 3. prioritize them by impact and frequency to cut through the noise so you can see the biggest opportunities clearly 4. connect them to your existing strategic roadmap so when the work gets done, we can automatically follow back up with your customers and win them over This is just step one of our mission to revolutionize how work gets done -- starting with what we choose to work on, and hope you follow us along for the ride!
Active Founders
Brian Kim
Brian Kim
Founder
Co-founder and CEO - Building Kestral - the future of work in the AI era After studying Computer Engineering at USC, I was an early engineer at GoFundMe, started my own company in corporate sustainability, and most recently led AI Workflow initiatives at Asana. Let's talk about human/AI collaboration, for-profit for-good business models, and endurance sports (USA Nationals qualifying triathlete). Also rap in a band and like dancing at parties/weddings.
Bernard Xie
Bernard Xie
Founder
Current founding CTO at Kestral, the AI native platform for work management. Previously at Asana, 2 time founding engineer (AllStripes: acquired by PicnicHealth (YC19), Terrastruct). On the side I golf, play the drums, and cook.
Company Launches
Kestral - Your AI Work Prioritizer
See original launch post

Hey YC, we’re Brian and Bernard, and we’re building Kestral, your AI Work Prioritizer!

TLDR; We make it clear what matters most to your customers and your business to guarantee your work has impact. (how? gotta read more)

https://youtu.be/R8Q5qOWRv2I

❌ The Problem 😟

Companies spend too much money on Sales and R&D teams just for them to 🥤suck🥤 at communicating with each other and this costs millions of dollars in blocked deals, churned customers, and time wasted on irrelevant roadmaps.

There’s gold in customer data in all its forms, yet even the most revered companies we talked to commonly lose sight of blockers and spend hours updating spreadsheets of feature requests.

uploaded image

📣 “We have deep anxiety if we’re deploying our product teams appropriately…” - C-Suite @ a Public Company
📣 “All my calls… in Gong. Feature requests … in Notion. We’ve got Salesforce. Why, in 2025, am I still grepping around for feature requests like a caveman?” - Head of Sales @ a Series B Startup
📣 “… we’re tracking feature requests on spreadsheets” - GTM @ a Major LLM Provider

Companies KNOW they should do better… but without the right tools, they are:

  • Getting outmaneuvered by competitors by failing to adapt to customer trends as they come up
  • Wasting hours manually tracking and escalating tasks that end up sitting in a backlog
  • Worrying if users will care about their “strategic investments”.

💭 📖 The Solution

uploaded image

We instantly bridge this divide between customer-facing roles and product development by:

1. Extracting ALL actionable tasks from a variety of sources (transcripts, Slack/Discord messages, GitHub issues…)

2. Consolidating duplicate asks and prioritizing them by theme

3. (optionally syncing to project management platforms like Linear/Notion until we displace them…)

4. Closing the loop with them, so you go from feature launch to immediate customer engagement & guaranteed impact

We’re building teams of AI agents that:

🤖 - understand the context of roles, projects, companies, and customers

🤖 - rank work on various heuristics like opportunity size, effort, and team goals

🤖 - make direct connections between customer asks and R&D’s work

🤖 - personally follow up with users or revitalize the pipeline to guarantee a positive business impact of work

Our technology can also easily transition to save orgs countless hours by eliminating work about work, like updating the state of tasks from meeting transcripts/messages and giving status updates while reprioritizing.

🤜💥🤛 The Team

uploaded image

(pre-YC vacation together in Jeju)

We’re close friends since college and multiple-time founding members of prior startups. We recently crossed paths at Asana, where Brian was an AI Tech Lead, while Bernie led Enterprise Authorization projects.

We’re leveraging our experience to combine the cutting edge of memory and knowledge base management with AI Agent Permissions for Enterprise to build the future of work.

We fully understand that work management structures are crucial to providing the necessary context for AI to deliver real enterprise value, and we’re not scared of any existing players in this space.

💡 The Opportunity

Our mandate is to disrupt the work management industry that has a $50B TAM, but we’re starting first by truly understanding and prioritizing what a company works on. As agents get better and commoditized, we aim to be best positioned to help companies get the work done faster.

The existing tools in this space can charge 5-6 figures a year for speculative and limited insights from customers at best, and they don’t even make the connection to R&D’s real work. We can instantly help validate/derisk costly roadmap plans, as well as help companies adapt to their customers’ needs immediately, providing millions in value.

We’re also personally seeing some very large investments (millions) in hiring teams to build internal systems to try to connect impact to work that is still manual at heart. Our partner said that if you’re seeing someone build something painful just for themselves… you’re on to something.

🟠 CALLING THE YC COMMUNITY 🧡

We’d love to meet more of you!
🗣️ In a CRO, Sales, GTM, or Customer Success role, and want your requests prioritized?
🗣️ Product Leader wanting to validate your strategy and get measurable results?
🗣️ Building the latest in AI knowledge bases and memory?

Let’s talk!

Would appreciate any warm intros to the above as well! We’re looking for partners to grow with us to become the most efficient, most actually customer-centric companies in the world.

Hit us up at founders@kestral.team or book a demo call, even if you’re just a little curious at https://kestral.team

uploaded image

Kestral
Founded:2025
Batch:Spring 2025
Team Size:3
Status:
Active
Location:San Francisco
Primary Partner:Andrew Miklas