Prayers Connect – Discover Iqamah times: simple idea to a full-fledged product

During my days when I used to drive to Maryland on Sunday evening to work at SparkPost and return to New York on Thursday after-work (due to Friday work-from-home policy), I faced one problem of discovering mosques around me to perform my prayers in congregation. It was more difficult for the fridays I stayed there. As friday is a work day, I can’t just go and wait long time for friday khutba (sermon) and prayers; I need to plan optimizely to go just in time of khutba and return immediately after prayer aiming to do everything from office-mosque-office in less than an hour.

But because I have no way to know exact timing, I often time went to early or too late. Also after work I would love to discover mosque and nearby halal foods places. Again it was tough to go to an area, check every website, many (if not most) also do not have one, find their website to get the iqamah times. Also some do not include updated iqamah times even if they have a website. So even though mosque were easy to find (using google map search), it was evidently difficult to know their iqamah times, services, any upcoming events and so on.

I talked to few other people and found many faced similar problems. Also a lot of NYC taxi drivers expressed their problems as they are always on the move. Plus during season changes, jama’ah (congregation) times change frequently (some weekly) and it’s tough to keep track of it.

Initially, I’ve tried to contact some of the existing services (islamic apps/websites) to see if I can collaborate with them and work with them to improve their platforms. However, I did not receive even a single response. So I decided to solve this problem myself.

Initially, I targeted New York, collected “all” mosques from different sources. I’ve developed a platform where I can manage these lists, modify them (today this panel is known as API – but it has a UI for internal users too) that powers website, mobile apps). Eventually, I hired a couple of data miner/editors from Bangladesh. They’ve worked hard to curate the lists, collect as much as possible from every available source.

I’ve then paired with a few other local friends who showed interest to work in this project. Among them, Hasan (Shakil) and Rahat were most notable and they’re actively contributing till today, alhamdulillah. May Allah reward them.

At first we allowed people to suggest iqamah times and that’s what we used to show on our site. But soon we found they become outdated pretty soon, sometime in a week. There is no point in showing outdated info. So we gave emphasis on collecting their iqamah schedules. Some mosques has monthly/yearly iqamah schedules which we added in our system. So they were good but not enough as most of the mosques do not have any fixed schedule.

To handle this problem, we’ve developed an algorithm to predict the iqamah times for a mosque. Today’s version of the algorithm is much more versatile than what we initially started building. Now we can fine tune it with some configuration for each mosque resulting more accurate results.

One important aspect of our algorithm is that while it’s not 100% correct, our goal is to ensure maximum chance of catching congregation. Hence, we took defensive approach in prediction and may predict a few mins earlier than actual iqamah time. So you may have to wait a few mins but we want to make sure you don’t miss the prayer due to our prediction was too late. However, we any place we show iqamah time, we also show a disclaimer marking whether it’s a prediction or verified time.

As soon as computer algorithm started helping us, we immediately expanded our listing to whole USA. After releasing Prayers Connect both iOS and Android apps, we thought we’ll create a website, mostly for an app landing page. However, we found we’re getting a lot of traffic on fridays. So we’ve also prioritized website just like our mobile apps. Now, alhamdulillah, results are good and the website is attracting most traffic. However, clearly, we’ll have to go very long way as we’ve so many dreams.

We’ve some more cool idea to improve the algorithm in the near future. However, we’re currently giving utmost priority to collecting a fixed schedule for all mosques. We’re working actively with mosque administration/imam to generate a yearly schedule and we’ve developed some tools to facilitate this. We’ve also added volunteering mechanisms so that anyone can contribute about the mosque they regularly go. We’ve added many many cool features both in the web and in mobile app.

Even though initially I thought this is a simple problem, eventually I learned it has many unique problems. As I kept working on it after my works, I’ve learned so many things about Ruby on Rails, React/Javascript, DevOps, React/React Native/Expo, iOS, Android etc. that I never learned through my job. I found out, it helps me do my job better as I’ve been learning other things too. I hope to get the rewards from Almighty Allah, but there are so many instant benefits in this project, alhamdulillah.

Checkout www.prayersconnect.com. We still have a very long way to go. If you would like to contribute to this project, please contact us, jazakallah.