I could use some inspiration. I feel “stuck”.
What should be my main focus for the next 6 months?
On one hand, I can think of quite a lot of things that either need to be rewritten or could be improved, and a few new features that would be super fun to create. But on the other hand, I often create new features without anyone really asking for them, and when done, no one really wants to use them.
So, I’m asking the community to spend an hour this Friday to share some insights with me.
What are people using OpenFlow for?
What are people really using openflow for ?
What features are missing or should be improved?
What features do you not know about? (Okay, maybe hard to answer, but I often experience people asking for a feature that already exists.)
What features is never going to be used ( like ML Studio that I’m working on right now )
Is it viable?
But also, I cannot pay my bills. Does it even make sense for me to keep working on this? Is there a need for this platform?
I tried donations, hosting of the platform, reselling consulting services, offering support plans, going open core, and offering licenses. Nothing seems to stick. Is it time to just give up and let it die?
I have 1 person donating 5 USD a month (big shoutout to Ivan Gordeyev for supporting me for multiple years). There were 4 more at one point, but most “gave up” after a few months to a year.
OpenIAP ApS has had 54 paying customers in total over the last 5 years.
OpenIAP ApS has 16 paying customers right now.
Those 16 customers generate a total of 4000 USD a month.
OpenIAP has on average over the year expenses of 2000 USD a month and pays on average 2100 USD in salary to me. ( that is 50% of what i would get on public wealthfare in denmark and is NOT enough to pay my bills )
To quote one of the many investors I talked to over the years:
“That is not a business, that is a successful hobby”
We must expand the community , other communication channels need
to be added : linkedin , youtube ,…
And we must focus on third world markets like Africa who do not have the financial resources to acquire paid solutions like uipath and create branches in each ountry.
BACCMED
I agree with you about expande and focus in Africa, I’m from Cabo Verde (West Africa) and where RPA is not used for most of the companies.
And I see a great opportunity to start using OpenIAP products.
Hi to all
Hi @allan
I participated in the meeting but I noticed that there were not many participants:
I think this may be due to the commitments of the contributors
I believe we need to create an open discussion (like chat) where we can share our ideas.
otherwise what are the decisions taken during this meeting?
hey
I used to have a chat ( first slack and after trying a few products ended up using rocket chat for a long time )
chat is really good when you need a quick back and forth dialog. Very convenient when troubleshooting or trying to get more details about a question. But it also requires more precens. If no one replies within a short period of time, new people will think it’s dead/empty and quickly move on. And it’s to stress full for me having to be online and present at all times. So to better fit expectation i moved to a forum. I start every day with spending 15 minutes to 1 hour, replying to all posts and then i safely ignore it until i have spare time again.
Discourse also supports chat, but I disabled it for the above reason. I don’t mind test enabling it for a short while to see if that would make the community more active and/or if people prefer that.
I was not able to be the part of the meeting on Friday, even if I was willing to, so I’m not sure what was the proposals and conclusions from it.
There is maybe one opportunity that came up to my mind about the possibilities of your products. It should be going forward in the way of AI and LLMs. I’m currently exploring the CrewAI that is using the abilities of LLMs Agents to work together on some tasks and provide final results to the user. Here your OpenFlow and his Agents and packages could be incorporated with this super cool tool, and also It could also be some kind of UI for tracking status (successful, failed) and other metrics about their work and results, and also managing those crew’s from CrewAI.
But for the income part, I’m not sure how. Except for the consultations about the tools and implementing them in companies or developing OpenFlow - CrewAI features based on their requirements, and giving the community opportunity to give feedback and create more examples of use cases and etc.
Also it is good to create some #hashtags for the #MadeWithOpenIapFlow or something like that to give more examples of the uses of your products, so the community interested in them and You also can get some feedback information about uses and their cases in real life.
Hope this gives you more ideas to think about it in this way!
Happy to help!
hi @Dusan_Perkin thank you for the feedback, and reminding me i needed to give feedback on the forum as well.
8 unique users joined the meeting, but one was 3 people so i guess 11 in total.
I had no agenda, except i had planned to show what i have been working on since november, and to have an open hearted discussion if i should continue developing this in an open source setting.
I was surprised to find peoples main inserted was about OpenRPA. I have always assumed OpenRPA was the “gateway drug” people took to then discovery OpenFlow and agent’s and then slowly stopped using OpenRPA.
But guess that makes sense, when I only have a commercial model around OpenFlow, there for that is what i’m primarily talking to people about.
Several people recommended more videos and guides to help people understand how they can “tie the different parts of the platform” together.
So this sparked a dialog about the future of OpenRPA and i have been thinking alot about that since the meeting. I can bring that up on the next meeting, if people want.
Next i tried showing what i have been working on, As all good demos nothing was working, and there was absolutely no interested from anyone in that .
For those not there. I have gone down 2 routes the last 3-4 months. I have developed a low code designer that allows building tensorflow.js layer models, and I have built a framework for scaling and distributing jupyter labs.
ml studio was intended to be for non data analysts to easily get started with using machine learning models with their data. Like categories, predictions etc.
jupyter stuff, was intended to be for people that already is working with AI, to allow them to continue in the tool they know (and maybe love?) and allow them to bring that directly into productions. By having both the built, debug and deployment tool in the same environment it should speed up the devops process.
next we talked a little bit about the future and commercial stuff. Nothing concrete came up, but people suggested creating more videos and guides in hopes that would bring more people toward the platform.
The atmosphere was pretty good, and I suggested we could do a meeting every friday or biweekly. people was not interested in weekly, but maybe biweekly, so i will try and set up a new meeting again for friday the 22th of march.
I think i found the issue i was talking about in the meeting.
I would love if someone would test if this version is working stable for them.
If i no one complains or find issues, i will make this latest release
Having experienced a variety of RPA products, I see OpenRPA as a promising free solution that has the potential to carve out its niche, especially among mid-sized companies within my network. My enthusiasm for learning OpenRPA continues unabated, fueled by the hope that this remarkable product will continue evolving and achieving greater heights. My gratitude goes to Allan for his support; I remain hopeful that this excellent product will persist in its growth and development.
At present, I am contemplating how to present OpenRPA to my clients and the means of installing it on their computers in an offline mode. In pursuit of this, I have explored the option of renting a server to schedule OpenRPA tasks and monitor processes, a strategy I intend to revisit. The cornerstone of this endeavor, undoubtedly, is the ongoing development support for this outstanding product. I am optimistic that the backing for this wonderful product will persist, encouraging developers like myself to further the product’s adoption and usage
Maybe I’m old school, but I prefer the forum format for support.
I’m a newbie to RPA and OpenRPA, so take my comments with grain of a salt.
I’m not a programmer, just an average computer user. I see a lot of potential for OpenRPA for individuals such as myself. However, the platform can be confusing if your background is not programming. For example, after several hours of watching YouTube tutorials, I’m not sure what when I need just OpenRPA, or if I also need NodeRed and OpenFlow. Then there is OpenIAP - not sure if that is even the same platform, but it references OpenRPA. Simple quick videos showing the differences and intended uses would be great (perhaps they exist and I haven’t found them yet).
Perhaps I am not the target audience, but my suggestions would be to expand the audience. Instead of coding and adding more functionality, provide more basic videos to lay basic foundation for non-programmers. Show real world examples where users save time(money).
Perhaps offer a service in which a user needing to solve one or two workflows pays to have a robot developed, paying an hourly fee or a job quote. For example, I have already invested several hours learning OpenRPA, but initially, I only have a few use cases for it. It would make more sense for individuals like me (non-programmers) to pay someone to write a robot than learn a new platform. I imagine you could develop the robot I need in just a few minutes. I know there are consultants, but their prices seem unreasonable, except for large businesses ($$$$/day!) Perhaps once those paying customers see the platform potential, they will want to learn it or just keep paying to have robots developed.
In summary, it seems providing more educational resources (dumbed down for newbies like myself) would expand the community and may be more beneficial at this time than adding more features.
Hopefully the platform sticks around. Allan, thank-you for your all of the work you have put into it!
I’ve uploaded several hundred hours of youtube videos.
I’ve spent many weeks with BPA to create the first documentation and many, many hours since then creating documentation, and writing guides, creating example repositories etc.
I have tried being active on linkedin, facebook and tiktok. Both Facebook and tiktok has never gotten any traction, so is pretty much dead, but linkedin seem to get some attention.
I don’t think adding more channels will help. We need someone who is better at grabbing people’s attention to do this, I’m clearly not skilled in that.
This is exactly why i wanted to make the platform(s) open source. If people who cannot afford the big commercial tools can get access to something that is enterprise ready for free, they can help by contributing back to the community. Offering support on the forum, expanding the documentation, testing features, sharing use cases, sharing knowledge about the platform on different media etc.
Yeah, i noticed that too. Same for second meeting.
I had the same idea. I’ve spent a long time working on a low code solution for building AI models called ML studio. I emailed all my paying customers asking if anyone was interested in trying it out, most did not reply, some asked for a demo, no one was interested in trying it out. The main arguments was either “we have decided to go with microsoft’s offerings” or " I don’t get it, what can I use it for?". Both super valid arguments, but that leaves me back at square one. What is the point in developing something if there is no need for it? If it does not fix a pain someone has and that they are willing to pay for.
I added ChatGPT plugins’ when that came out last summer. I added an OpenAPI interface to openflow when OpenAI launched GPT’s. I added “llm chat” to openflow to allow for “GPT like features” with open source models from within OpenFlow. I’ve added scalable Jupyter labs to Openflow, for easy distribution of Jupyter books. I created a framework for scaling out machine learning models.
Nothing seems to “catch” peoples interested. Either I’m not good enough at selling the solution or people don’t want AI. Either way, I’m the tool make, not a sales guy. It would be nice if someone else would help me bring the message out to more people.
Thank you. But people should really be thanking my wife. For the last 7 years, I have been unable to pay my bills, while trying to find a way to make a living of this, without “selling out” and crippling the product so it only works if you pay.
The intention with OpenRPA was to be a probe of NodeRED. You develop in NodeRED and use API’s as much as possible, and only if no other solution was possible you could fall back to an RPA solution. You can interface with multiple RPA solutions, but OpenRPA was made to make one that is better integrated and open source.
But there was a big demand from people that did not want to bother with maintaining both an openflow installation and openrpa, so “offline” mode was introduced. This is only intended for people that don’t need anything else than pure RPA features and have more complex scheduling demands.
OpenRPA: a robot designed to be RPA and only RPA. Unlike all other rpa vendors that combine API integration and RPA in the same tool
OpenFlow: A bad name, i need to rename it. Apache owns that name. My idea was to just rename it flow, or OpenIAP flow or maybe just OpenIAP ?
OpenIAP Aps: Is the name of my company, that owns the products.
IAP is an abbreviation of Intelligent Automation Platform
But you are right, the are missing a good overview somewhere. I’m just not sure where to put it ?
With OpenFlow 1.5 I moved away from “low code” only and introduced the agent concept. Agents allow writing real code, in your favorite developing tools, and easily deploy and scale you code across multiple platforms.
rpa: OpenRPA for low code users. robotframework, tagui, puppeteer, beautiful Soup and many others for coders.
api: Node-RED, elsa and more, for low code users. NodeJS/Typescript, C#, python/Jupyter, PowerShell for coders.
distribution: streamlined process for deploying code across all platforms ( docker/kubernetes for “running in cloud” , demon’s to run as background services on any server/client, and assistant for running in end user’s context.
I already, and have always offered consulting services. If someone need help with anything, they can purchase that on an per hourly basis from me. I do not build solution for people, but I have a list of partner’s that offer that as a service, many of those are also listed on my website.
This sentiment captures the essence of why I called for the meeting.
My theory was, maybe 1% to 3% of users of an open source platform can be converted into paying customers. With that logic, making it more user friendly would only help with that.
But so far, i have failed miserably on that part. I have payed a lot of money ( that I don’t have ) to hire people to help, i have spent an obscene amount of hours creating tons of videos, guides, documentation, example repositories and yet people keep complaining. I need help and I simply cannot afford to pay anyone to do it anymore.
And from an entrepreneur stand point, it makes no sense. I should find a pain people are willing to pay money to get fixed, and then focus only on what people are willing to pay for.
From day one, i decided I would make OpenFlow my “cach cow” and let OpenRPA be completely open source. Therefore there is no commercial model around OpenRPA. That also means any time i spent on OpenRPA is “wasted” unless it helps push potential customers to OpenFlow. I have seen what happens to projects that suddenly start removing features or crippling them, to force people to pay. I do not want to go down that route, but it is also not sustainable, that my wife has 90% of the financial burden, and I’m a one man army that is expected to be a jack of all trades.
If the community want to “step up” and help improving the usability, and broaden the user base then I’m also willing keep dedicating 100% of my time on developing the platforms. But if I’m still, after 7 years, is stuck as an one man army i will most likely start working on other, non open source tools moving forward, and then let OpenRPA die. That was the core reason for me calling for a meeting. To see if there was support for either lifting OpenRPA/OpenFlow out to a broader audience, or help find better financial model so I can keep working on the platforms.
I really liked the talks in the last two meetings, so I don’t mind continue holding them as a “free support” session, but I never got clarity on the big decision, regarding what i should focus my time on moving forward. I will continue working on that, and come to a conclusion soon i hope.
I’m head deep in re-writing a lot of code, so kind of forgot
But if you have something you would like to talk about, i’m happy to quickly setup a google meet ?
.
