I built an AI company to save my open source project

Geoffrey, thanks for sharing your story with us. OR sure is a weird niche. Good enough algorithms for solving these problems have existed for decades1, but we still see low adoption and your 95% estimate of the companies not optimizing their operations rings true.<p>Similarly to you, I spent a short while trying to sell VRP optimization with an API business model, and what dawned on me was that most companies do not have the necessary in-house expertise to integrate optimization into their existing tools even if the API is well-designed. There also really seems not to be any urgency to do that and most logistics companies just offload their inefficiencies onto their customers. Your routes are not effective? No problem, just bill more.<p>Some years ago I heard about a Swedish team of optimization experts who got so fed up with selling optimization to unwilling transportation companies that they founded their own—just to mop the floor with their ineffective competition. :D<p>I agree that ease of use is key here. In my PhD dissertation, I tried to address the issue by adding self-adaptivity within transportation management systems, mostly through automatic parameter tuning and algorithm selection. Such approaches remove some amount of fiddling when the optimization tool is adapted to a new optimization problem. Worth a look, perhaps, if you&#x27;re interested.<p>Many thanks again for the interesting article and all the best with Timefold.<p>1) E.g., already by the &#x27;90s, we had quite capable algorithms for the VRP. I have open-sourced a library of classical VRP algorithms called VeRyPy, containing simple and not-so-simple heuristic algorithms. It has enjoyed modest success among VRP researchers and practitioners. Nowhere near the success of OptaPlanner, but also, the purpose is different—OptaPlanner is production-ready, whereas VeRyPy is more geared towards education and research purposes.

I love that GOFAI companies are also calling themselves AI, because seriously why not. If a pile of linear algebra (fittingly trained on The Pile) can be AI, then so can optimizers and solvers.

I haven’t tried Timefold but did try a ton of solvers (local and web) a few years ago when trying to optimize MRP schedule. One of the hardest parts was converting my business logic into constraints, especially date based calculations.<p>Instead of explicit constraints, is there a way to provide a calculation that can be minmaxed? If every order has a due date, can I say +&#x2F;- 3 days = 0, 7 early = 9999 (not allowed), 7+ days late = (days late)^2?<p>Please email me (in profile) if you want to discuss.

It&#x27;s a constraint solver. You can find the community version at <a href="https:&#x2F;&#x2F;github.com&#x2F;TimefoldAI&#x2F;timefold-solver">https:&#x2F;&#x2F;github.com&#x2F;TimefoldAI&#x2F;timefold-solver</a>

Thank you for the hard work in this space! I think it is really important that there is a proper open source solution available.<p>I just found OptaPlanner and subsequently TimeFold few months ago, as I was searching for a solution for my wife&#x27;s veterinary clinics employee scheduling problem. The problem is not big enough for anyone to pay for the solution, but big enough to cause stress for whom ever is dealing with manually doings the shifts.<p>It was interesting that there were a lot of online SaaS providers that claim to solve the problem but they just simply are not configurable for all kinds of constraints of a real workplace.<p>Unfortunately I also feel partially same with TimeFold, because designing those constraints really requires changing the way of thinking of many problems. While the engine is capable of doing what ever, there is a steep learning curve to do it.<p>So while the article mentions documentation, I would say that the documentation is far from sufficient for wide adaption.<p>Personally, I would have really needed documentation about a mental model of thinking about the problem, and then a ton of examples how to solve real employee scheduling problems. Problem written in a format which the business people use and then translated into an elegant constraint rule explained step by step.<p>I had to invest more than 40 hours to get a working MVP that solves real problems, not just those that are already coded in the example code. Most people are not willing to do that.<p>What I&#x27;m trying to say is that to making planner software popular, it should be also usable for trivial projects. I understand that it&#x27;s hard to focus on everything, but just providing more information about real use cases and how they were solved and how to think about the design problems would make the market bigger, and bring you a lot more customers in the long run.<p>I just wonder how I might contribute to improve the documentation. I probably don&#x27;t have deep enough understanding of the correct solution, but I will look into it.

I just want to say that in 2024 there are _a_lot_ of hospitals doing shift planning manually and burning people out.<p>Also irrespective of the project itself, amazing wife!

Interesting story but what I don&#x27;t get is this<p>&gt; Founders might get a wage to avoid personal financial stress<p>How do you make a living as a founder if you don&#x27;t get a wage? Is that really a norm?<p>&gt; but then also need skin in the game, typically by investing personally<p>And also this. Does it assume that you will invest your savings or what exactly?

Great story, and great choices made.<p>The lesson in this to young people: you can either sell your time by the hour, or buy other&#x27;s time and use that productively. The former is very finite, the latter scales.

Cool to see Hawkeye 360 is one of their users. Hawkeye is an Earth observation company that builds satellites to monitor RF spectrum usage on the ground, so I can see how better tasking would help.

Inspiring story!<p>My favorite business course in undergrad was production operations management. Some time later I realized that in data engineering, critical path analysis is useful for optimizing directed acyclic graphs.<p>For example, it doesn&#x27;t make sense to optimize the longest running task when it&#x27;s not on the critical path. If you are optimizing for $, sure. But for time, there are better tasks to optimize.

Whats the name of the algorithm&#x2F;concept used here? I used simulated annealing, old fashioned algorithm for a college timetable prototype and I&#x27;m wondering how this company is using an AI model.

Good job ge0ffrey!<p>It is a tough niche market where you compete with excel people who have built their careers on their manual planning expertise.<p>Unfortunately you are pitching not to the end user but to their manager (who will have to force them to drop their spreadsheet mambo jumbo). But I guess this is the case with all AI products nowadays anyways.

Hmmm... Some co-founders and I are building a UK startup that needs to do skills-based task routing and optimize job assignment for workers who will drive to remote jobs.<p>Would love to have a quick chat with you or someone at timefold.ai to see if this is something we could potentially use.

Excellent article. You certainly managed to synthesize the startup experience with Open-source (two very challenging sectors). Do you just gravitate towards mathematically wicked problems? :)<p>Like you experienced, a good product without a team and without revenue is simply not enough, no matter how good the idea. I have unfortunately been one of the 90% of startup failures to experience that brutal reality.<p>My next project with a few others is to make open source coding &amp; development profitable. We want to keep advancing open source while also helping others to get paid for their efforts. We are still technically in stealth mode, but plan to be launching soon. Our website will be codebounty.ai (we don&#x27;t yet have it fully built out yet). With that being said, if you are interested to participate as a sponsor, maintainer, or contributor we would love to hear from you about what you would like to see in this program: <a href="https:&#x2F;&#x2F;docs.google.com&#x2F;forms&#x2F;d&#x2F;e&#x2F;1FAIpQLSdmlxJHIqOlwr1Nw3q0QwFpUowc6qFcuSNoybZz3rC5bi7_8g&#x2F;viewform?usp=header" rel="nofollow">https:&#x2F;&#x2F;docs.google.com&#x2F;forms&#x2F;d&#x2F;e&#x2F;1FAIpQLSdmlxJHIqOlwr1Nw3q0...</a> .<p>We would like to see it become a place where new coders can learn, develop skills, and get mentored by others. We would also like to have it as a way for others to have an income while looking for full time employment (rather than driving for uber). Heck, it could even become a recruiting tool for companies.<p>If you have ideas, I would love to hear from you so that we can help to make more amazing open source projecets like Timefold a more common reality!

What a great story and indeed you&#x27;re an excellent storyteller.<p>I want to wish you all the luck in the world with this, it sounds like it&#x27;s going well.<p>Have you considered offering public pricing for your API, or maybe a few trial? Even if it&#x27;s only for small use cases, I&#x27;d love to know I can just give you guys $100&#x2F;month to cover my small use case.

Thanks for your work and dedication on this problem domain, Geoffrey! I’ve used optaplanner in the past on several projects and will certainly use Timefold for future projects. This sort of constrain solver is an invaluable tool when you find yourself in the problem space and definitely worth the learning curve.

I need to check this out. In a past life my primary role was scheduling a call center of up to two hundred seats (at our peak). I had built out a complicated network of spreadsheets that used historical data and future projections to mostly automate the process, so much so that I was able to move us from what was monthly schedule releases to quarterly and had a full on preference survey that each agent filled out. It still was a few days work going through the responses and actually compiling the final result but what once was a headache for my manager I turned into 10% if my role and was able to focus on delivering value elsewhere by streamlining data reporting.

Honestly, I&#x27;m still not sure how I feel about Red Hat&#x2F;IBM&#x27;s journey. On one hand, RH killed (i.e. &quot;gave to IBM to rot away&quot;) a lot of products I liked (recently, the whole JBoss stack they had, see [1] for more details).<p>On the other hand, a lot of the products that RH gave to IBM were really painful to use. Let me preface this by saying that I mean no disrespect to Geoffrey and his amazing work! And, also, I suspected I was too dumb to understand OptaPlanner in particular.<p>But OptaPlanner was definitely amazingly painful to use. The docs seemed horrible to me at the time, and the UX of the product was on the level of &quot;the source code is there, to solve your problem, just understand the code base...&quot; To model your problems meant that ideally, you&#x27;d pay for Red Hat&#x27;s consulting, which would give you indirect access to the dev team, which would help you put your problem model into OptaPlanner.<p>A lot of the JBoss products I really liked were like that. Now, they&#x27;ll die by slow death over at IBM with nobody giving a damn. I guess such is life.<p>I wish Geoffrey all the best with Timefold! I truly hope you&#x27;ll turn it into whatever you desire, and that you&#x27;ll always have more customers knocking at your door than you can handle!<p>[1] <a href="https:&#x2F;&#x2F;www.redhat.com&#x2F;en&#x2F;blog&#x2F;evolving-our-middleware-strategy" rel="nofollow">https:&#x2F;&#x2F;www.redhat.com&#x2F;en&#x2F;blog&#x2F;evolving-our-middleware-strat...</a>

I read the entire article once I recognized the name.<p>Geoffrey is synonymous with OR&#x2F;Constraints solvers in my mind. I once had to tackle an issue in this domain and wound up using OptaPlanner after weeks of research.<p>It had the most intuitive API and the best community support by far.<p>It&#x27;s a bummer to hear that OptaPlanner was killed, but it sounds like ultimately it led to something greater, so there&#x27;s a silver lining.<p>Wishing you the best in your future endeavors, Geoffrey!

Nice project!<p>I am wondering if there is a risk, as it is open source, that a company with bigger marketing spend starts offering timefold in the cloud, kind like AWS with Elasticsearch? Or is there a big enough moat with the models on your cloud platform to prevent this situation?

Hi, I loved your story! I also checked out the app, but it&#x27;s giving me an error:<p>An error occurred in the Server Components render. The specific message is omitted in production builds to avoid leaking sensitive details. A digest property is included on this error instance which may provide additional details about the nature of the error.

I didn&#x27;t have much appreciation for shift planning until an engineering genius showed me the beautifully crafted schedule for their company (after I helped mess it up with some 12 hour days). They later went independent, with shift-planning software being one of their products.

I remember you talking about the project in the early drools IRC days. Your positivity and enthusiasm were what really made this project happen, and I have no doubts about your success with this new venture.<p>I’m glad things went well for you and that you managed to regain control of your life’s work.

Geoffrey is one of the best open source citizens.<p>Congrats on all this great progress! Please take care :-)

Thanks for sharing your company story!<p>I have kind of an off-topic question — how did y&#x27;all get a license for using SF Pro as a Web Font on your website? Afaik Apple on licenses the SF Pro as a font to only be used on Apple devices and apps?

&gt; Especially if there are almost no bugs, support is worthless.<p>This is why I don&#x27;t buy support from companies.

Tried their SaaS the day they launched and it was definitely very interesting

I&#x27;ve looked into optaplanner&#x2F;timefold but never ended up using or experimenting with it. Can anyone compare the experience to OR tools?

Congratulations and thanks to share your history, is inspiration for all of us who had an old project struggling to keep going

Awesome story, congrats on pursuing your passion

Eloquently explained, writing have have loads of wisdom earned and clubbed with benefit of hindsight.

Nice play with the name (Timefold).

Very enjoyable read. Hope you do post that follow-up in the future. Congratulations!

It is interesting how the combination of AI and open-source can provide new opportunities for project development. It is impressive to see it in action

All the best in your start-up

Great story, congrats!

Love to see a story with AI solving a real problem without generating a bunch of bullshit along the way

really interesting read!

awesome story!

wow

Great story and interesting product!<p>Reminds me of NextMv [1] loved their episode on SWE daily. Can anyone compare them to this and how they’re doing?<p>[1]: <a href="https:&#x2F;&#x2F;www.nextmv.io&#x2F;">https:&#x2F;&#x2F;www.nextmv.io&#x2F;</a>