Sabitlenmiş Tweet
Pokee AI
1.1K posts

Pokee AI
@Pokee_AI
Frontier Agent Running In Your Infra and Compute. Pokee AI's official X account.
Katılım Kasım 2024
11 Takip Edilen6.9K Takipçiler

4: Pokee-Match - Charlie, Joshua Logan, & Ken Goyarola (@Magicology, @Kaygeeartworks)
Pokee-Match took the familiar language of matchmaking and swiping and applied it to its own app experience.
It was a playful but very relevant example of how quickly builders can take an interaction pattern people already understand and rework it into a new product flow.
Simple, recognizable, and smart in the way it translated a proven UX into something new. Plus, their demo video is amazing! Check it out here.
English

We recently held a hackathon in San Francisco, and we are still so blown away by our winners! In particular, our 3rd place winners featured four very different projects, all super fun and with genuinely strong builds.
From audio tooling to marketplace agents to matchmaking apps and market analysis, this group really showed the range of what can be built fast with Pokee! We can’t recommend following them enough, and are excited to showcase their projects below!

English

The end result: a polished slides deck and presentation built around your goals, your style, and your content — without needing to do every step manually.
Just tell Pokee what you want the presentation to look and feel like, where to find the materials, and how deep to go!
It makes presentation-building way easier. Try it today at pokee.ai!
English

If the deck needs more depth, you can also have Pokee use the deep research skill (just ask it to use Deep Research) to make it more comprehensive.
That means it can help gather background information from various sources, sharpen the narrative, expand supporting points, and make sure the presentation has real substance behind it — not just surface-level slides.
So whether you need a fast pitch deck or a more researched presentation, the process stays very streamlined and easy!

English

Did you know that you can use Pokee to create a full slides presentation surprisingly fast? Here's how!
Just give it:
your media
your branding
your topic and
the look/feel you want
Alternatively, if you do not have everything ready, you can also ask Pokee to pull media and branding from the websites and sources you specify!
Then Pokee can take it from there.

English

And, if you want to get started, the fastest way to start using Pokee is simply:
Sign up, launch PokeeClaw, and ask for the outcome you want. It's that simple!
You do not need to know how to code, and there’s no long onboarding flow to sit through.
Just start chatting, and from there, simply dictate the tools and integrations you want to incorporate into your workflow!
Try it today! Pokee.ai
English

And once you’re in, you can ask for real cross-app workflows like:
“Pull bug-related GitHub issues, turn the critical ones into Jira tickets, write the summary in Google Docs, and post it to Slack.”
That kind of multi-app execution is exactly what Pokee is designed for.
It can also generate images, video, music, PDFs, presentations, spreadsheets, research, and code all from the same workspace!
English

Pokee is built for the way people actually want to automate work.
Just tell it what you want done in plain English and let it handle the workflow.
Slack. Shopify. Jira. GitHub. Gmail. Drive. Docs. Sheets. Slides. X. LinkedIn. YouTube. Notion. Asana. Dropbox. HubSpot. Image/video/audio generation and more.
Pokee currently supports 100+ integrations and growing!

English

Ever wanted to know how you might use PokeeClaw to create an automated training pipeline for LTX 2.3 LoRAs?
Look no further: We've prepared an article for you all explaining exactly how you can achieve these results with Pokee!
Pokee AI@Pokee_AI
English

I built a new workflow on @Pokee_AI called "Gmail AI Assistant: Inbox & Reply Automation."
This is what the agent does:
It takes the following inputs:
The Gmail address to fetch emails from.
A comma-separated list of categories to classify emails into.
The number of latest unread emails to fetch.
The email address to send the final report to.
The desired tone for the AI-generated replies.
It performs these tasks:
1. Fetches the specified number of latest unread emails from the provided email address.
2. Analyzes each email’s content to classify it, determine its urgency, generate a summary, and extract action items. If the email is important, it retrieves the conversation thread and generates a context-aware draft reply using the specified reply tone.
3. Compiles the analyzed information into a "Daily Briefing" and the generated replies into a "Suggested Replies" list, then composes a new email body containing both.
4. Sends the composed email to the receiver's email address.
It returns the following output:
The status of the sent email report (e.g., "Sent successfully").
I configured it to send me daily briefings of my inbox unprompted. I am yet to see if this works because I launched it today.
It has five input fields:
1. Reply tone
2. User email address
3. Email categories
4. Number of emails
5. Receiver's email address
For the email categories and reply tone, you need to be detailed and very specific with your prompt in order for the agent to fetch the right emails from your inbox.
I just sent a daily briefing to @Kyrannio and Nino; unfortunately, I wasn't specific enough in my email categories. It fetched the wrong information instead of "PokeeClaw product improvement feedback."
You can check out the workflow with the link below. Feel free to improve on it as you see fit. Pokee has a lot of third-party integrations, so make it wild and powerful!
pokee.ai/en/share-agent…

English

@Rufus87078959 @ZheqingZhu @CBinsights Thanks Rufus, we are so thrilled to have you all in the community alongside us!
English

@Pokee_AI @ZheqingZhu @CBinsights 👏👏, I'm not surprised. I expect Pokee to dominate the AI automation industry in the near future
English

Pokee AI named to @CBinsights' AI 100, the most promising 100 AI startups worldwide!
The bet: the next era of enterprise AI runs inside the security boundary, not outside it.
RL-trained agents. Long-context reasoning. Private cloud, on-prem, on-device.
Just getting started. 🚀

English
