Launch HN: TeamOut (YC W22) – AI agent for planning company retreats (app.teamout.com)
Here’s a demo: https://www.youtube.com/watch?v=QVyc-x-isjI. The product is live at https://app.teamout.com/ai and does not require signup.
We went through YC in 2022 but did not launch on HN at the time. Back then, the product was more traditional, closer to an Airbnb-style search marketplace. Over the past two years, after helping organize more than 1,200 events, we rebuilt the core system around an agent architecture that directly manages the planning process. With this new version live, it felt like the right moment to share it here since it represents a fundamentally different approach to planning events.
The problem: Planning a company retreat usually means choosing between three imperfect options: (1) Hire an event planner and pay significant fees and venue markups; (2) Do it yourself and spend dozens of hours on research, emails, and negotiation; or (3) Use tools like Airbnb that are not designed for group logistics or meeting space.
The difficulty is not just finding a venue. Even for 30 to 50 people, planning turns into weeks of back-and-forth emails for quotes, comparing inconsistent pricing across PDFs, and tracking budgets in spreadsheets. It becomes an ongoing coordination problem with evolving constraints and slow, asynchronous vendor responses. Most existing software is form-driven, but the real workflow is conversational and stateful.
Offsites are expensive and high stakes. A single event can represent a significant chunk of a team’s annual budget, and mistakes show up directly as cost overruns or poor experiences. Founders and operators often end up spending time on event logistics instead of their actual work.
I ran into this while organizing retreats at a previous company. Before TeamOut, I worked as an AI researcher at IBM on NLP and machine learning systems. Sitting inside long email threads and cost spreadsheets, it did not look like a marketplace gap to me. It looked like a reasoning and state management problem. As large language models improved at multi-step reasoning and tool use, it became realistic to automate the coordination layer itself.
Our Solution: The core agent relies on a combination of models such as Gemini, Claude, and GPT. A central LLM-based agent maintains planning context across turns and decides which specialized tool to call next. Each tool has a specific responsibility: - Venue search and filtering - Cost estimations (accommodation + flights) - Budget comparisons - Quote and outreach flows - Communication tool with our team
For venue recommendations across more than 10,000 venues, we do not rely purely on the language model. We embed both user requirements and venues into vector representations and retrieve candidates using similarity search. Hard constraints such as capacity and dates are applied first, and results are ranked before being presented.
On the interface side, we use a split layout: conversation on the left and structured results on the right. As you refine the plan in chat, the event updates in real time, allowing an iterative workflow rather than a static search experience.
What is different is that we treat event planning as a stateful coordination problem rather than a one-shot search query. The agent orchestrates tools, manages evolving constraints, and surfaces trade-offs explicitly. It does not invent venues or fabricate pricing, and it is not designed to replace human planners for very large or highly ...
30 comments
[ 2.5 ms ] story [ 55.6 ms ] threadIt’d be cool to offer one-off event suggestions, but I understand that’s probably not as easily monetizable.
Booking.com and similar moving into this space with their own generic AI tool.
Or even Gemini improving their UI so it presents search results more neatly.
Haven't organized large meetups, but for regular enterprise companies this could be a difficult to buy decision, because you have ChatGPT + bunch of connectors which can get company policies.
This could be good idea for event companies who regularly schedule things, but even for them, probably difficult to justify the value when you have access to ChatGPT and other connectors
Huh this surprised me as a forgone opportunity.
I heard second-hand about the process for organizing our last offsite. Searching for venues was not the time-consuming part.
The time-consuming part was actually engaging with the venues to confirm specific details not available online. Our teammate who did this engaged with _hundreds_ of venues. It was a lot of work on their part ... and probably not the most fun part of their job.
That seems like an ideal agent scenario?
Another challenge is travel, e.g: scheduling an event in Europe for a distributed team of U.S. people during bad weather leads to people stranded at airports, missing the event.
I think this is a great idea, but I am surprised to learn that organizers are spending most of their time communicating with hundreds of venues. Once you have a location and budget, finding a venue is straightforward.
> (2022)
Has there been a rebrand as of late? What was the product pitch before that? I guess "AI for planning company retreats" (and possibly SaaS for company retreats before that)
This capacity to pivot into these buzzwords shows that at least sometimes they are more phenomenons with marketing (or at least UX) definitions rather than technological ones.
"I want to have a two day offsite for a team of 12 in Cambridge in April."
It then started pulling up results in Cambridge UK. I meant Massachusetts. I didn't say that in the prompt, but I figured since there are two equally famous Cambridges, it would ask me for clarification.
I redid it specifying Massachusetts and it worked pretty well (although all the options it found were about double the price of what we actually booked).
An interesting idea!
BTW I didn't continue, but I assume you manage the whole booking process? How do deal with questions from the venue and other human in the loop issues?
1. Hoofddorp, Noord-Holland, Netherlands (actually ok location) 2. Marysville, Ohio, United States 3. Lisboa, Portugal 4. Nashville, Tennessee, United States 5. Kenmore, Washington, United States 6. Golden, Colorado, United States
I would expect there to be some reviewer agent that ensures that all found locations are at least within the same country?
But the quality of the actual AI response is just worse than GPT 5.2. Which makes it feel like a tacked on thing and more of a gpt wrapper.
I asked about a retreat with our US team that could also include one engineer in Pakistan that needs a visa. And the response was something to the effect of: "Assuming your engineer has a US visa, you can go to Puerto Rico".
Whereas chatgpt gave a much more well researched answer.
I completed a (rather large) contract to reverse-engineer, and eventually rebuild, a hotel chain's property management system from scratch from 2015-2018. We did it all: keycard integration, booking channel sync, credit cards, group bookings, yield management, front-desk GUI, supply management, taking rooms into/out of service, reservation migration from old system to new...you name it, we probably touched it. Dozens of small lessons about the lodging (and broader hospitality i.e. restaurants, country clubs, bars) business domain.
One thing is that hotel = brand (flag) + real estate + operations. You can remix those things in a lot of different ways, e.g. a single ownership group might have two properties on opposite sides of a street, one Hyatt the other Hilton, and they might look different but share staff, or procurement.
The industry's term for brand -- "flag" -- says a lot about how they view Hilton/Hyatt. They come and go, even if the staff running the property stays the same. The main reason hotels choose to flag vs. stay independent, is access to the chain's booking flow.
One of the more interesting consequences of this setup, is that small, independent hotels, are kind of a shit show in terms of technology. Chains generally require a lot of standardization of their member properties, including what software they run to manage the property. Many properties that don't affiliate with a chain don't have any property management system at all. It's basically 10-20 rooms run directly off the moral equivalent of an Excel sheet at the front desk. And why wouldn't it be--small boutique hotels often gross $1-2 million/year; there isn't budget for expensive enterprise software, or maybe more critically, the people who know how to deploy and operate it.
A significant value-add of Expedia and booking.com, especially with independent properties, is getting the supply (hotel) side of the market organized. Many of these hotels outsource their entire reservation tracking system to a single channel (e.g. booking.com) because trying to keep track of bookings across phone, direct web, Expedia, booking.com, and others, is just too hard without specialist software that requires more IT muscle to deploy than a single non-chain hotel can muster.
I mention this because I go to church every Sunday and was thinking about how much real estate churches have (event halls) that sit unused, and what a schlep it would be -- although good for everyone -- to expose the collective supply of the world's churches, HOAs, park districts, and other nonprofits, to the kind of events you're trying to do. It would indeed be a tremendous pain in the ass to get all the physical access (keys), contract terms, payment systems, availability, etc ironed out, but it's a massively underused class of real estate and many of these organizations could really use the cash.
I 100% agree, the interface is the least interesting part. Anyone can build a chat UI. That’s not the moat.
What matters is the messy stuff in the backend. Vendors. Hotels. Quotes that change. Someone forgetting to update availability. Contracts. Deposits. Random edge cases. That coordination layer is the real product.
The UI is just the tip. The hard part is keeping state across dozens of moving pieces and async back-and-forth.
My belief is that AI finally makes some of this operational glue automatable. Not in a magical way. But in a very practical loop:
ask → get info → update plan → trigger action → wait → adjust → repeat
Planning is basically structured ping-pong. It’s not search. It’s evolving constraints over time. That’s why it feels agent-shaped.
What I am basically saying is : Event planning is really AI agent prone and very conversation prone, that is why this kind of interface will take over travel and event planning. It is like you have a personal travel agent sitting next to you and it showing you options.
Totally agree with you though. organizing fragmented supply is the hard, unsexy, painful work. That’s where the real value is built.