DEV Community: Hack In The North

Getting prepped with Machine Learning skills for a hackathon.

Hack In The North — Tue, 11 Feb 2020 14:13:23 +0000

Introduction

Gone are the days when tons of lines of code had to be written just to make a simple neural network. With the rising interests of tech enthusiasts in the field of Artificial Intelligence, a plethora of libraries and tools are now available to get your machine learning model up and running in no time (well...let’s skip the training part, that does take a lot of time xD). Recent developments in machine learning libraries and tools have made building a machine learning model so easy that one can find open-source platforms like GitHub and tech blogs overflowing with different projects.

Tweaking with the model architecture and hyperparameters is one thing and actually applying machine learning knowledge to a hackathon is a totally different domain. This is, where I believe, most of the machine learning enthusiasts get stuck while approaching a hackathon problem. Well, if you are one who faces the same issue, this blog is for you!
In the coming sections, I have discussed everything starting from “how to arrive upon an idea”, to deploying your machine learning model on the web or an android app.

How to arrive upon a 'Machine Learning Idea" for a Hackathon

In recent times, mostly you will find two types of hackathons, one where you build a solution for a listed theme/ problem statement or other where you develop a solution to a problem statement that is closely aligned to their business. In either of the cases, one must be a jack of all trades and at least a king of one. A machine learning practitioner generally finds no difficulty in building up a model but may lack visualization skills and application skills.

One may not be familiar with the frontend and backend frameworks and thus it becomes hard for them to see their work as a complete product, which a hackathon expects. This is where it comes to choosing a team of people having complementing skills so that you can learn from one another and together come up with solutions that can be built into a complete product. From my hackathon experiences, if you want to generate some good hackathon ideas, your team must have at least one member from the frontend and backend domain so that they can transform your complex looking machine learning code into a product easily usable by the general public.

Being a machine learning practitioner myself, I will stage down the process into the following:

I am a Beginner, how shall I start:

Well, no worries, if you are a beginner. You shall be delighted to have chosen a field that would save humans from humans by serving one true purpose of destroying everything and everyone we ever loved, graciously letting humans walk hand in hand into extinction and bringing about a new world order governed by unbiased, sagacious and pure artificially intelligent robots( too soon?).

Anyway, for a newbie, it may be hard to come up with a project that uses machine learning at its core, but thanks to GitHub and the creative minds who push their projects on GitHub and make it available to the world, it’s a child’s play these days. I recommend exploring GitHub projects and getting an idea of the problems people have solved using machine learning. Going through different projects is an essential part of learning as it helps to know the limitations of different frameworks and helps you choose the one that best suits you and your team.

For example, The most commonly used frameworks in machine learning / deep learning are Tensorflow and Pytorch but it may not be possible to use either of them for every task. If you are developing an android app, you might wanna prefer Tensorflow over Pytorch for the sole reason that Google developed Tensorflow and has provided elegant tools to integrate Tensorflow code to the android app. You get the idea of different frameworks once you start exploring different projects on the internet.

I have made a couple of projects, how do I come up with a new one:

If you have made projects before, I am sure that by now you are already loyal to that one library, you consider, is the key to unlock the most appalling, petrifying yet necessary secrets of this universe so I’ll skip the part of explaining the pros and cons of different frameworks and will jump right away to the next part.

Well, if you have it in you to build a powerful model, you just need the right data to feed into your code. Websites like Kaggle, Analytics Vidya, HackerEarth and many more host regular competitions and make data available to the public. You may not participate in the competition directly but can exploit the data available on these websites to build a solution that solves a real-world problem. Looking at the datasets and other people’s code on Kaggle has always helped generations to come up with something new and that’s what I always do before finalizing a hackathon project.

Challenges with machine learning in hackathons

Machine learning models are no good lying in the IPy notebooks or scattered python scripts. To solve a real-world problem, the model needs to be deployed in a way it’s usable by the general public and not just by the domain experts. The main challenges you may face while creating an ML-driven product are :

Finding the right dataset:
While you may have the technical know-how on how to create the model in Keras/Py-Torch, but finding the right dataset which has the same features as your product demands could be tricky. Moreover, even if you find something which is somewhat intersecting with your needs, you might need to understand its structure and preprocess it to fit your specifications. Added to the limited time of a hackathon and the training time of your model, all this presents a big challenge.
Finding the right architecture:
Congratulations on getting your dataset in time, but are you confident that your run of the mill neural net will give you some presentable results? Did you try the Inception model? Is SSD fast enough for the demo? While there are some standard models for each domain you can try, but they all come with their pros, cons, and tradeoffs like good accuracy comes with slow speed, big size, etc. This is also one thing you might want to discuss with your team.
Source code:
If you are experienced in building models you might quickly get a model up and running and ready for the show but, If you are experienced in hackathons, however, you might not want to write it “during” the hackathon, especially from scratch. While “taking reference” from the internet is the default go-to for most people, the main challenge is to run it on your system. If the code is fairly new, chances are this may not be a problem, but if it is written in some other library, or on a deprecated version of your favorite library, you might have a hard time (CTRL+V)ing it.
Making it useful:
This is perhaps the biggest challenge when it comes to using machine learning in your project. The thing is, you have a great model (check), a well-processed data which represents your practical domain (wink wink, but check), the only problem is, you need its output somewhere off the realms of jupyter notebook, like on a website or more dangerously a mobile app. This is where you have to bring out the hacker in you. There is no default solution for this, you may have to deploy a local Flask server which acts as an ML backend, or host your model on Azure/AWS, which is costly, or use TensorFlow lite. Each of them has its own pros and cons. Be sure to think about this beforehand, since it requires some practice to get it running fluidly during a hackathon.

Perks of an ML project in Hackathons

Over the years, machine learning has found its way in every possible domain. You can find its use in almost every industry, be it healthcare, education, automation etc. That gives it the biggest advantage, you can opt for any track in the hackathon and look for a way to include machine learning which makes it all the more enchanting, natural yet seemingly impossible, making your hack more appealing to the judges and the users alike.

Coming back to Hackathons, we all must agree to one point that hackathons are all about presentation and being a machine learning practitioner does give you the power of finding relations among data and plotting complex looking graphs and charts. Analyzing a problem using past data and building a solution by exploiting it helps you build a concrete solution and make your hackathon pitches sound more confident and your end goal nearer.

With the advancements in computational power, deep learning has become a favored topic among researchers and open source enthusiasts. People have developed amazing libraries and tools which makes it very fast to get a machine learning model up and running. There are a lot of ready-to-use machine learning projects available on websites like Kaggle and Github which can be used directly or after a little tweaking for a hackathon project.

Having participated in a fair number of hackathons myself, I have seen the need of a machine learning practitioner in the team and have observed that almost every team has that one machine learning guy in their teams. Yeah, so being “the ML guy” in a hackathon does make you a core contributor to the project and you get to enjoy the hackathon doing a comparatively lesser amount of work (it’s the biggest perk according to me 😉).

Apart from the mythically formidable pheromones that “the ML guy” profuses, the person also enjoys the most creative space than the fellow hackathon colleagues. Many times, the judges are sold and swept off their feet for cheap owing to seemingly magical end results.

Technologies to learn

Here is a list of the best open-source AI technologies you can use to take your machine learning projects to the next level.

If you haven’t gotten your hands’ dirty writing machine learning codes, I suggest you to first get familiar with python tools like NumPy, pandas and matplotlib. Numpy is one of the most used libraries by any machine learning/ deep learning framework and mastering it will definitely help you get along with these frameworks. Pandas and matplotlib are mainly used for storing data and visualizing data respectively, you can use them for presentation purposes in hackathons.
Scikit-learn is an open-source library developed for machine learning. This traditional framework is written in Python and features several machine learning models including classification, regression, clustering, and dimensionality reduction. If you are not into building heavy deep learning models or want to develop a solution with less amount of data, you may use this library to build your solution.
Coming back to where we started, these days you just need a couple of lines of code to get your neural network running, all thanks to modern frameworks like TensorFlow, Keras and PyTorch. Keras is basically a high-level API that serves as a wrapper over TensorFlow (and theano) and saves you from horrifying mind-racking TensorFlow code. Keras is easy and intuitive and takes fairly lesser time to learn. Pytorch, again a high-level API developed by Facebook is an alternative to Google’s TensorFlow or Keras which serves the same purpose of building machine learning models quickly and optimally.
Well, if you are the one who doesn’t like high-level API’s and want to dive more into the core, TensorFlow 1.x/ TensorFlow 2.x is for you. With TensorFlow, you can create your own low-level functions, write parallel algorithms, build your own architecture from scratch and what not. Tensorflow is mainly used for research purposes and I personally don’t see its use in hackathons, given the ease of using Keras and the likes of it.

I am a Beginner, how shall I start:

I’ll say this again:

Hmmm, so here are some resources that have helped shape most of the modern hotshot ML Experts and you, my dear reader, will be next to avail this treasure.

Machine Learning by Andrew Ng: It’s one of the most popular courses and people usually start with this course to understand how machine learning works. The course is taught by Andrew Ng, an adjunct professor at Stanford University and a pioneer in machine learning and deep learning. The course has no prerequisite except basic mathematics and programming knowledge, it is meant for beginners and teaches everything from scratch. The only thing missing in the course content is some popularly used machine learning algorithms like Naive Bayes, Random Forest, Decision Trees, etc, but don’t worry, I have got all of these covered in the next course.
Course Link
Machine Learning A-Z: If you have gone through the machine learning course by Andrew Ng, you probably now want to see those algorithms working and build some applications using them. This course is mostly focussed on practical machine learning which unlike Andrew Ng’s course doesn’t dive deep into mathematics and focuses more on intuitive understanding and programming aspects. The course instructor, Kirill Eremenko gives a walkthrough to each line of code and makes you understand how to use different machine learning libraries. This course will leave you with one or two machine learning projects which you can put in your CV and can even use in hackathons.
Course Link
Deep Learning specialization: Coming to modern machine learning techniques, deep learning, this specialization takes you from the very scratch of neural networks to modern architectures like CNN, RNN, etc. The specialization consists of a set of 5 courses which may be taken individually or as a complete specialization, all of the courses are taught by Andrew Ng and have got very well structured programming assignments that gets you comfortable with Tensorflow and Keras. Along with the programming exercises, the courses also get into the depth of mathematics and the core working of the deep learning algorithms.
Course Link
Deep Learning A-Z: Don’t have the patience of completing those 5 long exhausting courses, no worries, try this course, from the very start of this course you will get your hands on the modern deep learning frameworks and will use them throughout the course to build some amazing applications. This course will teach you how to set up a deep learning environment in your own machine and will make you install important libraries and tools which help you throughout your machine learning journey.
Course Link
TensorFlow: Data and Deployment Specialization: Now that you are backed up with the theory and know-how to write machine learning architectures in Python, R, Matlab, etc. It’s time you put your skills to develop an actual application, which is readily usable by the general public. The specialization consists of a set of 4 courses, teaching you how to deploy your model on the browser, Android / IOS, raspberry pi and microcontrollers. You will get an understanding of how to use a suite of tools in TensorFlow to more effectively leverage data and train your model.
Course Link

Carefully choosing the right set of courses and having a continuous gradient flow in your head to learn this amazing technology will definitely bring out a machine learning expert in you and will help outperform others in hackathons.

About the authors

We are information technology undergraduates studying in IIITA. Feel free to contact us via the following channels:

Dhruv Agarwal
- Email
- Linkedin
Shivansh Beohar
- Email
- Linkedin

Talking to the veterans of the game: Team sudo code.

Hack In The North — Wed, 15 Jan 2020 15:47:25 +0000

Hailing from the same alma mater that proudly hosts Hack In The North every year, they have been a constant inspiration and a formidable force to be reckoned with in any hackathon. Considered to be the team that was promised in their alma mater, they have captured many forts throughout the country namely, Smart India Hackathon, Hack36, Topcoder Humblefool Hackathon, Prototype IIITA and yours truely, Hack In The North. Hyperbole aside, these humans of IIITA can easily charm Thanos into killing himself through their quirky, witty and polished answers to any posed question or circumstance. In a quick interview with the sudo code, they relive their experience as we slip into the nitty-gritty of hackathons.

(L-R) Himanshu Shekhar, Madhurjya Pegu, Abhishek Sharma, Souvik Sen in HACK36 2018.

Q. Hi, Thank you for taking the time to talk to us. Can you introduce your team members?

Abhishek: Very humbled to be sharing our little experiences with you. Thank you for having us. So we are a team of 4 - Abhishek, Souvik, Madhurjya, and Himanshu. We all went to IIIT Allahabad together (Batch of 2015) and spent a few semesters attending a lot of hackathons. During the process, we made some cool shit and a lot of memories and – all in all, had a great time (and some horrible nights :P).
Coming to the specific roles, generally, I handle the design and Madhurjya plays around with React to give life to my designs (and sometimes rewrite my ugly code). Himanshu takes care of the backend, the software architecture and integrations. Souvik’s project manager whose only job is to make sure that the team doesn’t implode.

Q. You guys have had the same team for hackathons for quite a long time. How much do you think bringing together a "right" team is important for a hackathon?

Abhishek: Really important. There are two aspects to building a team in my opinion. First, have people with a complementary skill set so that you can build an all-round product. Ideally, your team should have a designer, a front end dev, a backend dev and someone with good product knowledge and communication skills who can refine and pitch the idea. These could be different persons or a combination. For example. In our team, I am normally the designer and front end dev.

Secondly, I feel it’s also important to have people in your team who share the same skillset as you (or have at least some idea) so that whenever you’re stuck you can ask for help from your fellow team members. Himanshu and Madhurjya used to help me in frontend debugging (because I write shitty code). Also, Madhurjya used to give me design inputs.

Himanshu: Apart from “skillset” and “workflow”, we were close friends and have developed great trust and respect for each other which I feel is the ingredient of the right team. We always had an idea what one of us might need/be thinking about in any case and were nearly predictable in terms of how we will function during the hackathons. Over the years, the coordination was just obvious. We didn’t have to make the same push as the first hackathon.

As Abhishek usually says, “A team that stays together, slays together”, and we were there.

Souvik: If the intention is to just have a good time, I think it’s better to team up with strangers; but if winning something’s the objective, then obviously it’s much better to have people on your team who know different parts of the tech stack.

Madhurjya: A right team is made from quite a few factors such as different types of skill sets (design, frontend, backend), having open ears for one another. One of the important reasons why we made a good team was that we were good friends in and outside the hackathons.

Q. How does your team start prepping for a hackathon and in what timeline?

Abhishek: After registering for a hackathon, we always promise that we shall meet and brainstorm ideas. Although we do meet, 100% of the time we end up goofing around on silly ideas with no tangible outcomes. So it’s mostly on our way to the hackathon venue or during the first few hours when we decide what to build. The next few hours are mostly spent in speccing out the features and dividing the initial work amongst ourselves.

Himanshu: I’d still want to have those goofing-around-silly-discussions because they were really fun. Although there were no immediate outcomes, discarding a lot of ideas helped to figure out what stuff we care about building and what we don’t care about. We bonded well and tried to develop our own “authentic” taste about what is a cool idea to work on together.

Souvik: Zero prep every time. 😎

Madhurjya: We tried brainstorming a lot before the hackathons when we first started. Later on, we gave up since all of the brainstormings didn’t produce any conclusions. Sometimes, we started with some inspiration from sources like ProductHunt, TechCrunch, YourStory, etc. and In the end, went for ideas that we could relate to which ultimately helped us do well.

Q. What was the most treasured hackathon of your team? What was your strategy and idea for the project? Is there something you would do differently if you had to do it all over again?

Abhishek : For me, it has to be Hack in the North 2.0. It was my first-ever hackathon and the first one for sudo code. It was almost 3 years ago, so I don’t really recall the strategy, to be honest. We created a small utility called Moodmonk , which basically ran sentiment analysis on voice to detect your mood and suggest solutions. Coming to doing things differently, I would probably not choose Himanshu in our team. Just kidding :P

Himanshu: For me, HINT 3.0 was a great one. This was the time when Consensys was the title sponsor and we decided to make a blockchain-based app -Aerially.
The app from an implementation point of view floats smart contracts when you are denied boarding a plane and fulfills the contract when other airlines give you a seat.
The strategy was simple - Sharma + Pegu would take care of design + frontend, Souvik would take care of backend + interfacing + any viz required, I decided to implement the blockchain part.
The best part here was the learning curve from being clueless about how smart contract works to make it work flawlessly overnight. This was long after I was tired of writing ReST APIs and wanted to make something different.

In case I had to do something differently, I would have asked for help more often. Ashoka Finley (one of the judges) helped us a lot with technical challenges we didn’t know about. But I wasted a lot of time being hesitant.

Souvik: Any of the ones where Himanshu didn’t fart in people’s faces.
The strategy was that Himanshu wouldn’t fart - especially in someone’s face.
If we had to do something differently, I would ask Himanshu not to fart in anyone’s face.

Madhurjya: There were many moments I treasure, from traveling to eating together. Discussing ideas for the upcoming hackathon which leads to random truth and dare sessions :p.

Q. Can you describe some other projects and what you liked about them?

Himanshu: Via is another hackathon project I personally liked. We built it during HackInOut 2018, Bangalore. It was a web-app made for mobile which tracked the quality of the road using motion sensors on your phone.
For me, things that gave a feel about “Oh shit, this seems impossible to do” till the very end and then you work your ass off to make it happen (“surprise, it worked”) is a great experience.
This in recent times had been one of them where we had a tough time to figure out how we can judge road conditions from the motion sensor when the sensor has noise on its own. However, some fundamental concepts learned for exams - thresholding, high-pass filters, correlation theory - helped us figure it out.

Although we didn’t win here but felt accomplished from the experience + the trip experience we get from attending hackathons in another city.

Souvik: Any project where we had a clue of what the end product would look like was more fun. I think Dexter was the one where we were the most in control. Plus it helped that the rest of the projects at that hackathon sucked, so we won. 🙂

Madhurjya: In my opinion, Aerially is one of the most unique ideas we came up with together.

Q. What is the difference between a Hackathon project and a production-ready product in your opinion?

Abhishek: I approach the hackathon projects from a demo point of view. You work on bare minimum features and try to put out a Proof of Concept. “If it’s good enough for a demo, ship it.” On the other hand, production-ready projects are very different in the sense that you need to carefully look at them from scalability and security point of view in addition to taking care of all possible use cases.

Himanshu: Both are way too different. This was a very tough lesson for me and was one of the reasons I stressed others out during the demo. Being someone who usually writes backend code, I always wanted things to be functional and never ever wanted to compromise on code-quality (at least for the code which passed by eyes). (The definition of quality changed over time).

However, I realized that hardly anyone cared if my method is elegant, or I have the best engineering practices in place. There were very very few cases when the judge was interested in the software architecture/software design part of the hack.

It’s certainly very important for a real-world product and cannot be compromised. However, your due diligence again depends on what stage the product is in - beta, early adoption (aiming fast iterations), large adoption (aiming stability and scaling).

TLDR; If the outcome isn’t working, none of it matters - be it a project or product.

Souvik: No difference. Production-ready from the start. Himanshu writes the tests. 😛

Madhurjya: Hackathons have a much shorter time frame compared to production-ready projects. In hackathon projects, there is no good file structure, no tests, etc.

Q. For many people, participating in their first hackathon can be intimidating. What would you tell those people?

Abhishek: You don’t need to know everything before participating in your first hackathon. Just take a leap of faith and jump right in. You don’t need to have a kickass idea or be a know it all person. If you are a couple of friends thinking to solve a small problem, that’s more than enough to enter a hackathon. Normally, there’s enough mentorship and time available in a hackathon to figure and do stuff on your own. That’s also the best way to learn stuff in my opinion, pick a language/framework, decide on a use case and just search around – “How do X in Y language” and repeat.

Himanshu: My first hackathon experience was HINT 1.0 (2016) (with Souvik, Dewan). To be honest, it was a disaster. The hack was a tamper-monkey script, which is basically a javascript running on console and I had zero knowledge about JavaScript. We spent an entire day to figure out XHR which is pretty naive and just a copy-paste away. It was Dewan who made things work in the last 6 hours.

Moreover, the demo didn’t present well (didn’t know how to extend display :P), the time-slot was over and my mic was cut-off right when I was speaking (time’s up!). Consider all happening at your first hackathon when you’re in the first year.

But that’s how it started. You didn’t run the first time you tried to walk.

Souvik: Have no expectations of winning, and go with the sole intention of getting as many stickers as you can.

Madhurjya: Hackathons can be quite useful for people who want to start software/product development. For the first few hackathons, I only learned how to use an API from 3rd party sources or make a simple one myself. To a beginner, it is one for the best to get into development.

Q. Lastly, how was your experience participating in HINT? What was your biggest takeaway?

Abhishek: Hack in the North will always hold a special place in my heart since it was my first ever hackathon. Over the different editions, I have learned a lot about creating products and met amazing people – Thanks to HINT. The biggest takeaway for me is to just focus on creating kickass stuff and having fun, everything else will automatically fall in place.

Himanshu: HINT has been one of the best experiences I had in my college so far. The biggest takeaway was meeting not just amazing, but truly great people and learning about how to do stuff IRL, with a team.
Kudos and thanks to Team HINT over the years for making it all happen.

Souvik: HINT has always been one of the more fun hackathons and it’s gotten better each year. The only thing I don’t like about it is that the acronym should be HITN.

Madhurjya: HINT has contributed a lot to my software dev journey. The organizers do an awesome job to keep the culture alive. Everyone should at least try to participate in HINT.

(R-L) Souvik Sen, Himanshu Shekhar, Madhurjya Pegu, Abhishek Sharma in Hack In The North 2018.

Getting prepped with Web Development skills for a hackathon.

Hack In The North — Tue, 07 Jan 2020 11:30:22 +0000

Introduction

Ever wondered what is the backbone of almost all the software-based hackathon projects? Ever wondered what is that one field which will allow you to contribute most towards making your hackathon idea into a fully realized product? The answer to these fundamental questions in web development. And guess what, it's one of the easiest avenues of development to get into!

In this blog, we will give you a complete rundown of how to start off with WebD and by the end make you prepared to tackle the stressful demands that any hackathon asks of its web developers.

We have divided this post into the following sections:

Where do I start? (Or "enough of the pep talk, give me some actual technical details!" )
Where do I study from? (Or "sources that'll enlighten me in the ways of web development")
What do I do now? (Or "I wanna escape this tutorial purgatory, suggest me some sweet project ideas to test my skills"*)
Hackathon experience (Or "what are the valuable hackathon experiences that you guys can share!?")
Tips and tricks (Or "I heard pragmatism is the name of the game, kindly impart some of your nifty tips and tricks")
About the author (Or "you have written a lot guys, got credibility?")

Last but not least, this post is an attempt to give you a big-picture idea of the perspective of web developers when they are participating in hackathons. We hope that you enjoy this read. Let us dive in!

Where do I start?

Now, the prerequisites to starting off can be really challenging, intimidating and might not be the first thing you want to hear. Do you have a web browser like Chrome or Firefox and a simple source-code editor, say all platform available VS Code? Cool, now that we are in clear, let’s move forward :P.

Web development can be divided into three broad practices, namely:

Back-end development - Also known as server-side development. It deals with all the behind-the-scenes logic of your web-app.
Front-end development - This deals with setting up the "look and feels" of the website. Basically, as a front-end developer, you'll be designing the UI of your web-app, ensuring that all the interactions with your web-app are as user-friendly as possible.
Full-stack development - This refers to building both the back-end and front-end parts of the web-app.

At this point, you'll have to ask yourself what kind of developer you want to be. If you want to work closely with the designer and make the best looking website out there, just learn front-end. If you want to deal with the core logic side of development, enabling maximum functionality and performance out of your web-app, just learn back-end.
Wanna be independent of the woes of searching for that dreamy complementary in skills, developer, who would make your mother proud and can perfectly grasp the depth of work you intend to do while reasonably compromising in the diversity of the project?
Then, full-stack it is.

For an in-depth look at how the modern web dev roadmap looks, check out this awesome video.

Where do I study this from?

There are a plethora of learning resources online. From full-blown online courses to useful youtube educators, here's a shortlist that you may want to pursue depending on what you want to learn.

Courses:

YouTube channels:

The only advice is that it's best to follow a learn-as-you-go approach because life is too short and your list of ideas and wishes, well let's face it, even made Santa climb up the chimney and take early retirement. So, start off with a Bootcamp style course, develop a strong intuition and immediately start making your own projects. That's it! As you face challenges in your projects, your intuition will guide you to the appropriate resources. At the end, when push comes to shove, that's what you'll be doing in hackathons!

What do I do now?

Now that you know what about things that constitute Web Development, let’s look at a few projects are you can work on as self-help exercises. Before that you must know that with the extent of knowledge people nowadays manage to acquire before even having any hands-on experience, these projects might seem trivial at first. And they pretty much are for an experienced developer. But for the beginners, nothing substitutes good practice. Let’s move forward now.

Front End

Make your website.
Obviously enough, to start being a ‘website developer’, you need to be a website developer. What better way to begin than creating your own personalized one. While its pure fun to add anything and everything in it that describes you, you can also tailor it to fulfill specific needs like, say your online portfolio or your own blog. It doesn’t have to be perfect but surely try to move in that direction. Use a version control system obviously. Make mistakes. Fix them. Look at real examples. It’s learning by doing.
Make a social media authentication page.
You have often signed up for/ logged in to social media accounts. The forms you fill are often fancy and conveniently spaced out. The colors on the page look good too. Try replicating that. Or better, design your own form, with layouts and colors of your choice. Try making it look as professional as the ones you’ve seen. This can teach you a lot about front-end design, from designing good layouts and deciding on good color combinations to actually place components in accordance with the layout perfectly.

Back End

Host your websites on a local server.
Now that you have your framework of choice, try reading its documentation and learn how to run it on your local server.
Connect to a database
Learn about a database management system like SQL or MongoDB. Try connecting your backend server to the database. Make the front end interact with the database. You might build a blogging website where the blogs that the user writes will be stored in the system.

Hackathon experience

We have been to some hackathons and have also helped in organizing one. The very draw of a hackathon is rapid development since generally, they run for 30-40 hours. And that makes it all the more difficult. And the key is a lot of practice. And experience. If you are interested in hackathons then you should participate in a lot of them. Learning comes automatically. Here’s a rundown of what goes through your head during a hackathon.

In the beginning, you will be more anxious about the quality of the idea than its actual development. You’d constantly be thinking about whether your idea is better than the other participants’. You will find others with ideas that you might find more interesting than yours and might even wonder if you should change it. It’s not generally wise to do that.
During development, you still wouldn’t be very worried about the clock. Until the first setback, which will definitely come. Especially for beginners, it is unwise to assume that there won’t be problems irrespective of however much you’ve thought it through. This is a phenomenon not exclusive to just hackathons. And now you might panic.
Hopefully, you don’t face too many setbacks. ‘Too many’ differs for different persons. In any case, once the number of setbacks reaches your threshold of self-motivation, you might start getting bored. You will start questioning the very idea of hackathons itself. You’ll wonder who came up with the stupid idea of developing good software in just 1-2 days. Wouldn’t the limited time compromise the quality of the software? Don’t dwell on it too long though.
The clock will soon strike completion time. You’re not at all close to completion. Your panic further increases. You start working faster. You start bodging things. Just remember, it’s perfectly fine to do so. Especially if you’re a beginner.
The hackathon has completed. Your project might or might not have been. You realize a lot of other people are in the same pain. Actually, most of the time you might not even realize it. The questions from point 3 flood your mind again. But make a point to never get disheartened. It’s a process even the most experienced hackathon participants go through in the beginning stages.

Tips and tricks

Points in the preceding topic must be seeming a bit disheartening. They present to you the problems that a participant generally(read always) faces in a hackathon. The most comforting thing I can say now is that it’s totally fine to mess up. The motive is to learn, not be a development rockstar from the get-go. Still, a few tips and tricks might prevent much anguish.

Don’t try to build a software/hardware aiming to solve a large problem that you might not be able to complete even in a week rather than during a hackathon. A lot of people fall into that trap.
Be prepared with your idea before the hackathon begins. While some hackathons might provide you with their ideas at the commencement, generally, you’re allowed to come up with your idea beforehand. So obviously come up with an idea beforehand.
Keep a calm mind throughout and trust your skills. Simple and obvious.
Be prepared to forgo some features of your proposed software/hardware due to lack of time. It happens to the best of them. Have a contingency plan for such situations. Follow good programming paradigms to ensure a missing feature doesn’t render the whole project non-functional.
Sometimes the UI is more important than the backend. While you shouldn’t put all your focus on UI without any substance to the backend, but good-looking projects are often graded better than the relatively uglier ones. Hence do make your product look as good as possible.

About the authors

We are information technology undergraduates studying in IIITA. Feel free to contact us via the following channels:

Arpit Mishra
- Email
- Linkedin
Vishal Pani
- Email
- Linkedin

Hack In The North 5 | IIIT Allahabad

Hack In The North — Sat, 04 Jan 2020 14:39:41 +0000

Greetings developers,

Hack In The North is the collaboration of programmers, designers and inventors to code, design and build innovative solutions that tackle real-world problems.

We are the core team organizers of Hack In The North 5, IIIT Allahabad to be held in March 2020. Hack In The North 5 is the fifth iteration of what started out as an effort from unorthodox students to allow student developers to have a platform where they can showcase their ideas and skills.

As we move closer to the commencement of the hackathon, we take it upon ourselves to bring you few words by hackathon veterans and domain experts in the hopes of starting a culture of learning and building through careful inspiration.

In the following weeks leading up to the hackathon, we will be publishing blogs to help the aspirers get a better broader picture of our shared goal (to make you a better developer xD).

For more details you can kindly check out the Hack In The North website :
www.hackinthenorth.com

Our Facebook page: https://www.facebook.com/HackInTheNorth/

To register a team, simply head to our website https://www.hackinthenorth.com and click on the “Apply with Devfolio” button - in case of questions, check out the FAQs and feel free to reach out to us on the page or via comments!

Email us your queries, we are happy to help: hackinthenorth@iiita.ac.in