Giselle
Build and run AI workflows. Open source.
980 followers
Build and run AI workflows. Open source.
980 followers
Built to design and run AI workflows that actually complete. Zero infra setup—just build and run. Handle complex, long-running tasks with a visual node editor and real-time tracking. Combine models from multiple providers in one canvas.
Giselle
Hello everyone! 🙌
I'm so excited to finally launch Giselle and share it with all of you!
We built this for ourselves. There are countless AI workflow builders out there—but when we actually tried to use them for real work, something always felt off. Too complex to set up, too rigid to adapt, or too opaque to debug when things went wrong.
So we built what we actually wanted to use:
A visual node editor where you can see your entire workflow at a glance
Mix and match models from different providers in one canvas
Real-time tracking so you know exactly what's happening
Zero infrastructure headaches—just build and run
It's open source because we believe the best tools grow with their community.
If you've ever felt frustrated with existing AI workflow tools, give Giselle a try. We'd love to hear what you think.
And if you know of better products out there, please let us know! We're always looking to learn from great tools.
@codenote Hey! transparency and debuggability are big differentiators. How do you handle versioning and reproducibility in visual workflows, so teams can confidently change nodes or models without breaking existing pipelines or losing the ability to trace past runs?
Giselle
@malekmoumtaz san, Thanks so much for the comment!
Versioning and reproducibility are exactly the areas we're looking to improve.
Currently, we do version the execution results of workflows, but we don't yet have versioning for the workflows themselves. Features like Figma/FigJam's auto-save history or the ability to manually save with named versions are definitely on our wishlist—and something we're actively working toward.
As you pointed out, giving teams that confidence and peace of mind is really important to us, so we're committed to making this happen.
Giselle
@malekmoumtaz Great point — versioning and reproducibility are absolutely critical once workflows move from “experiment” to “shared team asset.”
We’re not fully there yet, but it’s a priority for us to achieve both: the fast, intuitive iteration you get from a visual editor and the kind of robust version control teams need to make changes safely. The direction we’re heading is to make it easy to evolve a workflow (swap nodes/models, tweak prompts/params) while still being able to confidently refer back to what ran before and why it behaved the way it did.
On debuggability, we’re already collecting a lot of internal execution data, and we plan to expose that in-product in a user-friendly way — think clearer run history, inputs/outputs per node, and logs/metadata that let you trace what happened in past runs.
Giselle
@malekmoumtaz Thank you for this question. It really gets to the core of what teams need when transitioning from experimentation to production. We'll keep building on this — your feedback helps us prioritize.
@codenote Huge congrats. Upvoted!
This is a breath of fresh air for the AI automation space! I completely relate to that feeling of "whack-a-mole" when trying to debug a complex LLM chain—it’s usually so opaque that you have no idea where the logic broke.
The visual node editor is the real hero here. Being able to mix and match models from different providers on a single canvas is exactly the kind of flexibility power users need.
Question for the team: How does the real-time tracking handle high-concurrency workflows? Is there a specific "debug mode" that lets you inspect the exact prompt/response at each node?
🚀
Giselle
@annnarobbb Thank you so much!
We rely on @Trigger.dev as our workflow engine to handle this.
Internally, we have the capability through @Langfuse , but we haven't exposed it to users yet. A debug mode is actually something I want myself too, so I'm adding it to our roadmap.
Thanks for the great suggestion!
Giselle
@codenote @annnarobbb
Thank you so much for the kind words!
We don't have a dedicated "debug mode" in the strict sense yet, but in our Studio editor, you can configure and run prompts for each node individually. This gives you a pretty good idea of what's happening at each step. That said, we recognize that understanding how outputs chain together across nodes is really the key—and we're actively thinking about ways to make that flow more visible and intuitive.
Giselle
@aditya_shynxmedia san, Thank you — I'm glad the "actually finish" part resonated. That's exactly what we care about most.
You're right that tools like Giselle are better experienced than explained. We believe showing real value through actual workflows matters more than loud marketing.
Your approach to creator-led discovery is really interesting. Small experiments in real scenarios, not promos — that sounds like exactly the kind of feedback we're looking for.
Giselle
I'm Taka, CEO of the team behind Giselle. Today we're launching Giselle — a visual AI app builder designed for product teams.
Why we built this:
We started building Giselle over a year ago, back when GPT-3 was the standard and tools like CrewAI and n8n were just emerging. Our original goal was to bring LLM-powered automation to consulting and finance — domains drowning in research and documentation work.
But here's what we learned: getting "professional-grade" output quality was hard. Really hard. So we did what any stubborn team would do — we dogfooded relentlessly. We became our own zero-customer, using Giselle daily to build our own products.
That journey shaped what Giselle is today: an AI app builder optimized for product ops and GitHub-native workflows.
What makes Giselle different:
GitHub as your vector store — Turn your repos, issues, PRs, and code into RAG-ready context with one click. No pipeline setup.
Event-driven workflows — Trigger Giselle apps from GitHub events (new issue, PR comment, etc.). Build your own CodeRabbit-style review agent — no code required.
Team-first, cloud-native — Apps you build are instantly shareable. Call them from a chat UI ("Stage") or directly from GitHub with custom slash commands (you define the command name).
What you can build:
✅ Automated PR review agents
✅ PRD drafters that pull context from your codebase
✅ Spec/docs updaters triggered by merged PRs
✅ Parallel workflows like Cursor or Claude Code — but for your whole team
Why this matters:
Tools like Cursor and Claude Code have supercharged individual developers. But teams still struggle to share that leverage. Giselle bridges that gap — not everyone needs to be a builder; one person's app becomes the whole team's productivity boost.
What's next:
Right now we're focused on product ops, but the path forward is clear. As we expand support for diverse document types and data sources, we expect Giselle to handle the consulting and professional services use cases we originally envisioned — research synthesis, client deliverables, knowledge management at scale.
In the near term, we're doubling down on two fronts:
Visual builder improvements — Making it even easier to prototype AI apps without code
Developer-facing features — Instant API access for any app you build, MCP (Model Context Protocol) support, app virtualization for complex compositions, and smoother paths to scale with LangChain when you're ready
We'd love your feedback. What workflows would you build first?
Giselle
I wrote an article about GitHub-powered Vector Stores—one of Giselle's standout features. Very few tools make it this easy to turn your codebase into a Vector Store. Would love to hear your thoughts!
How to Use Vector Store and Query in Giselle
https://giselles.ai/blog/vector-store-and-query-in-giselle
Giselle
I've also written an article covering some more advanced use cases. Feel free to check it out for reference.
Cross-Repository Analysis with Giselle: An Advanced Use Case for Vector Store
https://giselles.ai/blog/cross-repository-analysis-with-giselle-vector-store
Giselle
Continuing the series!
This one covers how to build a custom deep research workflow—no prompt engineering expertise required.
Building a Simple but Powerful Custom Deep Research App with Giselle
https://giselles.ai/blog/custom-deep-research-app
Got a use case you'd like to see covered? Drop a comment—I'm always looking for ideas.
Giselle
Continuing the series! Last time I covered custom deep research workflows—this one's about bringing any document into your AI workflows.
Beyond Code: Building RAG Systems from Any Document with Giselle
https://giselles.ai/blog/document-vector-store-in-giselle
Most business knowledge isn't in code—it's in PDFs, Word docs, and internal wikis. Document Vector Store lets you upload these files and query them with AI, no coding required.
The tutorial walks through building a PostgreSQL documentation Q&A system using Context7's pre-processed docs.
Got questions about RAG or document ingestion? Happy to chat in the comments.
Giselle
New article! This time we're flipping the approach—instead of pulling knowledge from documents, we're reacting to GitHub events as they happen.
GitHub Event-Driven Workflows: Building Automated Issue Assistants with Giselle https://giselles.ai/blog/github-event-driven-workflows-giselle
The tutorial walks through building an issue assistant that automatically researches context and posts helpful comments when new issues are created. Set up takes about 15 minutes, no code required.
Got questions about event-driven automation? Drop them in the comments.
Giselle
Taking it further! Yesterday's issue assistant was just the warm-up—now we're building something teams actually ask for: a custom PR review agent.
Building Your Own PR Review Agent with Giselle
https://giselles.ai/blog/pr-review-agent-giselle
CodeRabbit and Copilot are great, but every team has different needs. This tutorial shows how to build a review agent that understands your codebase using Agentic RAG—so feedback is grounded in your actual code, not generic best practices.
Giselle
Hey Product Hunt 🙋♀️
We just launched Giselle — and I'm the designer behind it.
While our engineers were focused on making AI workflows safe to rely on — even when they run for hours — I obsessed over a different question:
What does it feel like to build one?
I spent way too much time on something most people won't consciously notice: the nodes. We designed them to feel like bright stars floating in space — vivid enough to stay readable as workflows grow, but calm enough not to overwhelm you.⭐️⭐️⭐️
My goal was simple: I wanted your workflows to look like something you'd actually want to show off — not just tolerate using.
I'd love to hear from you: What feels intuitive when you're building — and where do you still feel lost?✨
Giselle
Thanks again for all the feedback from the launch.
I published a short write-up on how we built Giselle using Giselle — sharing some of the workflows and decisions behind the scenes. 😉
In case it’s helpful:
→ Designing Giselle With using Giselle: Closing the gap between design and code
Giselle
@kaochannel154 Thanks for writing and sharing this post!
It was really interesting to get a glimpse into your thought process and see how you approached things from your perspective.
Giselle
@kaochannel154 san, Thank you for the lovely comment! I'm truly grateful that you brought the vision of "stars floating in space" to life.
As engineers, we tend to focus on reliability and stability, and the "experience of using" often takes a back seat. But because you obsessed over what it feels like to build, Giselle became not just a tool, but a product that's genuinely pleasant to use.
I'm so happy to be launching with this team. 🚀
Giselle
Appreciate the thoughtful feedback here. 🤩
We’re already testing a few workflow tweaks based on what people shared.
Giselle
As we wrap up this week, we just wanted to say thank you again for all the thoughtful conversations here.
We’ve been reading every comment and have already started incorporating some of the feedback into Giselle.
We also shared a behind-the-scenes write-up that focuses on the practical how — how we actually built and run AI workflows using Giselle itself:
→ https://giselles.ai/blog/document-vector-store-in-giselle
Grateful for this community 🙏
Product Hunt Wrapped 2025
Hits a nerve tbh. I’ve bounced off a bunch of workflow tools—too much setup, no clue when jobs hang. Open source + real-time view + mix providers sounds right. Gonna try it on a long-running scrape/summarize flow. Curious about retries/state.
Giselle
@alexcloudstar san! Thanks for the comment!
We've struggled with the same frustrations ourselves, so we're really happy to have built and launched a product that addresses them!
For long-running tasks, theoretically there's no time limit. That said, for scraping use cases specifically, we might be missing some features that could leave you wanting more—so if you have specific needs, we'd love to hear about them. You're also welcome to post an Idea here: https://github.com/giselles-ai/giselle/discussions/new/choose
On retries, we currently do a single retry, but we're thinking it'd be great to let users customize this themselves—definitely something we want to implement.
Giselle
@alexcloudstar san
Thanks so much for the comment - really appreciate it! Would love to hear how it goes with your scrape/summarize flow. Feel free to reach out if you have questions about retries/state management. And if you discover any interesting use cases, please share!✨
Giselle
@alexcloudstar Totally feel you — a lot of provider-specific “workflow” tools are convenient, but it’s frustrating when you can’t mix in other models/services. That was a big motivation for building Giselle as open-source + multi-provider from day one, with real-time visibility so you’re not guessing where/why a job is stuck.
And yes on retries/state: the execution side is basically ready — what we’re working through now is the UI/UX for it. We don’t want a “retry button somewhere” that’s technically there but awkward to use. We’re exploring how to surface retries at the most natural place/time in the flow (e.g., right on the node/step that failed, with enough context to decide whether to rerun/adjust).
Giselle
@alexcloudstar Thank you so much for the kind comment.
One of Giselle’s strengths is that it requires no infrastructure setup and is easy for anyone to get started with.
We’ve designed the UI to be as simple and intuitive as possible, so you can build workflows quickly without dealing with complex configuration.
We care a lot about making the experience approachable even for first-time users.
Giselle
@alexcloudstar This really resonates — we've bounced off plenty of tools for the same reasons. Hope Giselle works well for your scrape/summarize flow. Let us know how it goes!
DiffSense
I'm curiouse: What's the 3 coolest things that has been built with this?
Giselle
@conduit_design san! Thanks for asking!!
Our team's been building all sorts of cool stuff, but my personal favorites are:
Giselle's QA Assistant
An implementation planning app for Giselle
A blog post drafting app
Since Giselle is fully open source, you can actually see the QA Assistant in action here: https://github.com/giselles-ai/giselle/pull/2582#issuecomment-3695472517
Giselle
Love those examples, especially the QA Assistant.
It’s become an essential part of our development workflow, and we use it every day to review changes and maintain quality.
It’s been incredibly helpful for our own development process.
Giselle
@washizuryo Indeed! It's been a game-changer for our workflow too—glad the example connected with you.
Giselle
@conduit_design san
It's still early days, so I don't have three examples yet, but I can share a personal favorite - I used Giselle to build a stylish media site for my dog! It really showcased how versatile Giselle can be for passion projects. You can check it out here: https://giselles.ai/blog/making-stylish-dog-media-with-giselle
More cool projects are being built as we speak, so stay tuned!✨
Giselle
@conduit_design Thanks for the question.
Here are three things I personally use Giselle for most often:
Generating PR titles and descriptions from diffs
This is probably my most-used workflow. When I saw DiffSense, I immediately understood the value — generating structured context from diffs is incredibly practical. I also think the Apple Silicon–native, local-first approach you took is very elegant.
Design reviews from onboarding screenshots
I often capture screenshots of product onboarding flows and run design review workflows on them to identify UX issues, inconsistencies, or areas to refine. It’s been useful for iterating quickly and objectively.
Lightweight research to catch up on artists
As a more casual use case, I use it to collect and summarize recent updates about artists I like from multiple sources. It’s especially helpful for catching up on artists I haven’t followed closely in a while.
Looking forward to seeing more real-world use cases emerge.
Giselle
@conduit_design Thanks for the question — and your product is cool too!
Might overlap with what others have shared, but here are my personal favorites that I use daily:
PR review agent
PRD analysis agent
Team meeting agenda agent
Vuetify for Figma
@conduit_design @gyu07 I’m not getting what the QA assist does. The GitHub link doesn’t open due to many requests. :)
Giselle
@conduit_design @muhammad_satar QA Assist is an agent designed for people reviewing pull requests — it helps figure out what to review and in what order. Going through commit logs, comments, and code manually to determine what needs QA is a lot of work. Especially these days, with AI writing more code, human review is getting harder.
As for the GitHub connection — I hear you. We've tried to keep it simple with a standard auth flow for developer tools, but if you're running into issues, please don't hesitate to let us know!
Giselle
@muhammad_satar san, Thanks for your comment! If you need any help, feel free to reach out to our support team anytime.
Triforce Todos
I appreciate the open-source approach here. It makes it feel less like a black box and more like a tool I can actually trust and grow with.
Giselle
@abod_rehman san, Thanks so much!
That really means a lot to hear!! We're excited to keep growing the product while staying transparent and building that trust.
Giselle
@abod_rehman Thank you — I completely feel the same way.
When I see an impressive AI product, if it’s not open source, there’s always a small sense of unease. I often find myself thinking, “This feature is incredibly useful, but how is it actually implemented?” That curiosity is part of what draws me to OSS.
My own skills today are built on reading and contributing to the great open-source projects that came before me. Giselle is our attempt to give something back to that ecosystem — something people can trust, inspect, extend, and grow alongside. I’d be very happy if this project becomes a small part of that ongoing open-source tradition.
Giselle
@abod_rehman Thanks for this — it's exactly why we went open source. Trust is earned, not claimed. Glad it resonates!
Giselle
@abod_rehman Thanks, that means a lot to us.
We believe that being open and transparent is what builds real trust.
And we hope it also encourages people to help shape where this goes next.
Giselle
@abod_rehman Thank you, that means a lot.
We wanted Giselle to feel understandable and trustworthy, not like a black box — and open-source is a big part of that.
Appreciate you sharing this.
Giselle
@nico Thanks so much for checking it out, Nico! Really appreciate it 🙌
Curious — what part feels the coolest to you so far? For most people it’s either the visual node editor + real-time tracking, or the ability to mix models from different providers in one canvas.
Giselle
@nico Thank you so much for the kind words!
We’re really happy to hear your positive feedback, and we’ll keep working on improving the product.
Giselle
@nico san
Thanks so much—really appreciate it!
I’m super happy you think it’s cool. If you get a chance, I’d love to hear what stood out to you most✨
Giselle
@nico Thanks so much — really appreciate it! 🙌
Giselle
@nico san, Thank you so much! 🙌 We'd love to hear your thoughts if you get a chance to try Giselle!
@codenote I am actually thinking about integrating this with a fork of @Sokosumi and the Masumi protocol
Giselle
@nico Oh, that sounds great! Please let me know once you've completed the integration!!