Echoing Exciting Experiences

Categories Alexa, AWS, Lambda

So a few years ago I began experimenting with building skills on the amazon alexa platform. I found the developer experience to be top-notch and the sdk’s provided to be easy to use. I created two skills name phill and joe. From my understanding developing a skill on the Alexa platform consists of 3 basic components.


Intents are the VUI (voice user interface) equivalent of a software interface. They indicate the typical list of features and functionality to wish to accomplish from your Alexa Skill.

The above are the intents I had assigned to the skill Joe. Which are to perform sentiment analysis, Say a greeting, Send a text message, validate a Two-Factor Authentication code and perform a secure action.


Utterances are like the implementations of the above intents. They are essentially the product of applying context to intents to make them easier to understand and implement. Think of them as test cases for human interaction with our skill.

The above are utterances which basically train alexa to understand which intent to pickup when receiving a particular type of input.

The Actual Skill itself:

Alexa skill support a variety of execution backends. For ease of integration and convenience I chose to use AWS lambda. Attached below is the code in javascript for the skill joe.

I made use of different client libraries for all the various intents I wanted to accomplish. Such as twilio for sending text messages. speakeasy for two-fa and a simple sentiment analyzer. Lambda allows for configuring environment variables which contained all my configurations in a separate env file that could be uploaded directly to aws.

Architecture Diagram

Scope for improvement:

  • Try not to write a blog post about something you built 3 years ago. 😛
  • Setting up CI/CD to simplify the development process
  • Continue to have fun with whatever you are planning to achieve.

Add some Spark-le to your life.

Categories Apache Spark, NodeJS, Raspberry Pi, Scala

Hey folks, I know I have been under the radar for a really long time. Its because I have attempted a lot of different projects throughout 2018 and the later half of 2017. Some of the projects have not had the desired level of success. However there is no such thing as failure there are only lessons waiting to be learned.

With that being said this was a project I had done in the later half of 2017.  The idea was given to me by a former colleague. The goal was to scrape data from a bar’s website which shared the amount of a given brand remaining at any given time. After I have collected this data I wanted to perform some time series analysis using the ARIMA model and execute it on an Apache Spark cluster. Rough architecture diagram is attached below.

First implementation of the project

So the website I was trying to scrape had a widget that shows a small barrel which would change color / height of filling depending upon how much beer of that particular brand was left behind. I needed to parse some CSS so I found cheerio.js to be helpful in ingesting html and rendering the required elements correctly. Once that was done using some simple mathematics I was able to compute the percentage remaining. Attached the censored source code for the scraper below

I ran the above for about 2 weeks on one of my VPS’s. Once I had enough data I went ahead and began implementing the ARIMA model code which I got from here.

So initially it was taking forever to run the model on two raspberry pi’s plus my macbook, So I decided to add some more resources and scale up my solution. Using the brilliant scaleway cli I spun up 2 VC1M instances and added them as slaves in my deployment. That changed my architecture as follows.

Adding additional compute nodes as slaves

The great thing about the scaleway cli is that its commands resemble the docker cli ,which is in muscle memory for me. I could spin up/down vms on the fly just like I would do with docker containers. This made the project a lot more affordable as I could spin up the compute nodes only when I want to run the model. I had to do a small hack by running a local tunnel on my macbook so that the vms could call into my laptop. So now that all the sunshine and rainbows part is over lets get to the actual scrutiny of the project.

So why did this project fail ?

  • Data was very inconsistent.
  • CSS rendering was inaccurate leading to bad measurements.
  • Not enough samples were collected.
  • Database was not ideal for time-series data analysis

What were the key learnings of this experiment?

  • Apache Spark is an amazing distributed computing engine that can be setup easily and runs across different architectures ARM,x86 etc due to the portable nature of the JVM.
  • Data is the new gold, without good data you cannot do any kind of useful analysis.
  • We are living in an awesome timeline where we can request resources dynamically and pay only for what we use.
  • Always do some basic groundwork like a technical evaluation before starting any project.

Thats all for this blog post stay tuned for more exciting and successful projects coming up!

Disclaimer: I have tried not to mention any brands/locations or url’s for privacy reasons. Also contributions are welcome!

IoT (dash) Button

Categories AWS, IoT, Lambda

Hey everyone, I know it’s been a while I have been busy working on a bunch of exciting stuff to keep you guys entertained. This post is about a fun approach to solving a common problem we face. Wouldn’t it be nice if you could summon everyone at the press of a button. Well now you can using an AWS IoT Button. The original ones were given away for free at an AWS ReInvent event some time ago. The IoT button is basically a Dash button that is unlocked, programmable and 3 times more expensive.

Someone please explain this sorcery

Leaving all the business reasons aside. The IoT button comes with a non-removable battery and is good for only about 2000 clicks. Amazon definitely has everything figured out it took me just 5 minutes to get up and running. I just unboxed the button downloaded the app on my phone. Connected to the Wi-Fi, logged into my AWS account and next thing you know I could see all my lambda functions and API gateway endpoints that could be assigned to it. So the ideal use case for me was to invite everyone in the household for a given meal. It is normally a hassle to coordinate with everyone and try to work things out this project simplifies it down to a single click. This is what the solution architecture looks like.

I have chosen to use the service twilio to achieve the end goal of notifying users. Twilio is an amazingly affordable telecommunication service that provides a programmable interface. The provide both REST endpoints as well as libraries for popular languages like Python, Javascript etc. You will see more projects of mine using them in the future. So each IoT button comes with a GUID dcn number on the back of it. This dcn number can be using the register the Button as a Thing for AWS IoT. Once the button is registered It is available as a trigger that can invoke a variety of AWS services. I have decided that 2017 is going to be the year of serverless therefore I have built this app to be run serverless using AWS lambda. As with everything I do this project is completely open sourced and can be cloned at

Its actually really cool so when the button is clicked there is a small led indicator on the IoT button that blinks white when triggered and keeps blinking for a while. It eventually turns either green (success) or red (failure) depending upon the outcome of the Lambda execution. The result being a text message sent from my Twilio number which is as follows.

The size of this project is quite small but it has proven to be very useful. That being said there is always scope for improvement.

  • Add scheduling to determine who is availble on weekdays/weekends for which meal
  • Incorporate the ability to handle a response from to user to confirm/decline their availibility
  • Incase text message delivery fails have an alternate notification mechanism such as email

Hope you guys enjoyed this post. Expect more exciting posts to ECHO in the future. 😉


Categories AWS, Computer Vision, Docker, Photography, Raspberry Pi

Hey Everyone,
Sorry for the lack of updates, I have been working on something so awesome it should technically be 3 blog posts and not one. It was such an intense project that I ended up bricking one of my Raspberry PIs by corrupting the memory card and causing segmentation faults. The entire fiasco is also what slowed down my progress. But anyways to start off this new year I wanted to shift my focus on upcoming and bleeding edge technologies like OpenCV. The overall idea is to find the most dominant color in a given frame so that if something was to remain camouflaged it would have the best chances with the chosen color. To implement this I used K-means clustering to divide the image into two sections and determine which color occupied the most space. The efficiency of this algorithm improves as we increase the value of K (the number of clusters). But for the sake of speed I chose to use only 2 clusters. Here is what the algorithm looks like

  1. Capture video using RPI camera
  2. Stream the video as a supported format MJPEG
  3. Load the video into OpenCV
  4. Process every frame as a Numpy Array
  5. Reduce the size of the Image for easier computation
  6. Using K Means Cluster create a histogram with K sections
  7. Determine largest section in histogram
  8. Render color on 8×8 LED Grid

The solution architecture is as follows
Abra Kedabra!

At first, I tried to everything using only my 2 raspberry pi’s but the problems I face was that it took 14 hours to compile! and the performance was incredibly poor. So I thought it was best to delegate the responsibilities to a container in the cloud which was very easy to setup and configure. They are 3 main components in the system.

  1. MJPEG streamer (here)
  2. AWS EC2 CV instance
  3. REST API for the SenseHat (by yours truly)

Check it out in action.

So after installing the MJPEG streaming module on my Pi2 I wrote a simple wrapper shell script for it.

This would create a MJPEG stream at 'http://< rpi-ip >:8080/?action=stream'
The next step was to consume this stream in AWS. I created a simple base container using the anaconda framework for python. setting OpenCV was as easy as conda install opencv . Next is the meat of the project code for which is shared below.

So this is what the EC2 container sees.


And this is the histogram generated after K Means clustering.

As you can see Red seems to be the most dominant color in the frame. You can tell by the amount of time taken for the neural network to compute the dominant color that this project is in an infancy stage. Let me mention the scope for improvement for this project.

  1. It is fundamentally wrong to use a value of k=2, I need k to be the exact number of different colors
  2. To provide the color for the LED board I should use a pub-sub system instead of REST as acknowledgment of request is not necessary
  3. In order to achieve true camouflage only computing to colour is not enough I need to figure out patterns and textures
  4. Overall performance of the system must improve by using a distributed system approach like (MPI) or tweaking the algorithm

Hope you guys liked my project. Look forward to more bleeding edge projects in the year ahead

/gifsuggest slash-command

Categories AWS, Docker, Lambda, RealTimeMessaging, Slack

We have all heard of IaaS, SaaS and PaaS offerings. But I recently came across AWS Lambda, which I would like to define as RaaS (Runtime as a Service). So what AWS lambda provides is an execution environment for running NodeJS and Python code in a completely Serverless/Stateless manner. which means all we have to do is write code that conforms to the specs of a lambda function and let AWS handle the scaling and execution of the code. The good news being that it integrates well with many AWS services and you get billed only for the compute time you use. You can trigger a Lambda function in response to many events such as an S3 upload or a change in a Code Commit repository. Since I wanted to play around with this technology I thought of building a slash command that suggests gifs for a given term. The solution architecture for this application would be as follows.

many moving parts
many moving parts

So the sequence of events is as follows.

  1. When we type in /gifsuggest "something" slack makes a POST request to our app, providing us with lots of information such as the user, team, channel etc. They also provide a response_url whos purpose I will explain later.
  2. Using the nginx config on my VPS I redirect the request to a containerized express app.
  3. Slack user experience guidelines enforce a rule which states that a response must be made in 3000ms or else the command is considered a failure. Which is why at this time I just send some placeholder text immediately to slack.
  4. The next step is we POST the search term and the response url to an API Gateway endpoint.
  5. The API Gateway is what triggers the execution of our lambda function.
  6. The lambda function GETs gifs using the giphy api.
  7. Finally using the response_url from earlier we POST the gifs and create a slack message. Slack allows us to use the same response url to create 5 messages in half an hour.

Below is the slash command in action.

Slash command in action

Now let’s get into the code. There are many moving parts in this application so I will show each microservice in the order of which they execute. The first is the express app that quickly replies to slack and then starts the lambda function. I have called this service slack-lb.

This app is run in an alpine Linux docker container with a nodejs environment.
The next microservice is the Lambda function itself.
There are 2 main parts of invoking a lambda function in javascript.

  1. Event: This contains the data that is being passed into the invocation.
  2. Context: This contains the lifecycle information of the lambda function such as execution time remaining and other lifecycle hooks.

Amazon is nice enough to provide a NodeJS library that registers the AWS lambda context as an Express middleware. This makes migrating existing express apps to Lambda very easy. We just need to write a handler as follows.

and then register middleware in our express app.

To deploy the lambda function all we need to do is create a zip file with the node modules, the handler and the express app files and then upload them on AWS.

Benefits of Lamda:

  1. No need to manage any infrastructure.
  2. Automatic scaling.
  3. Easy monitoring using cloudwatch.

Issues with Lambda:

  1. Currently only supports NodeJS 4.3
  2. Only supports stateless applications
  3. No support for sending files

Scope for improving this project:

  1. Right now once gifs are sent the user still has to copy the url and paste it at the destination there should be a way to forward the gifs using message buttons.
  2. The app is using a public api key for giphy which is rate limited and not peak performance. Must get a production key
  3. Migrate the proxy slack-lb app into its own lambda function.

PS: A lot of people make this mistake but GIF stands for Graphical Interchangeable Format. which would mean that it is gif and not jiff.
P.P.S: Like everything I do this is opensource. Feel free to contribute