Projects
Collection of Projects I did as part of Course work or self motivation :)
YouTube Ads
Student Product Consultant for Measuring social course in CMU
Worked with the YouTube Ads team. The problem statement was to understand how YouTube Ads could power shopping for Gen-Z. The initial research involved a lot of data/research analysis to understand Gen-Z, what they value, and what their characteristics are. Demographic deep dive. After this, we conducted a survey that reached 1000+ individuals and got around 250 responses to get quantitative data on Gen-Z shopping habits on socials. We also worked on interviewing 12 Gen-Z candidates to understand their social media shopping patterns to get some useful user insights. We used these insights & quantitative data to power are recommendations to the YouTube Ads team.
​
Environmental Policy Innovation Center
Student Product Consultant for Policy Innovation Lab course in CMU
Worked with the EPIC team to rethink the feedback & community engagement within the Climate and Economic Justice Screening Tool made by the Council of Environmental Equality to support President Biden's Justice 40 initiative. Did 9+ user interviews of Academic researchers, advocacy groups & EJ activists. The core issue turned out to be a lack of bridge between the community members & the government - to share feedback about the tool, the data used in the tool, their representation on the tool. Decided to ideate on solutions to bridge this gap between the community members & the government by making the feedback & engagement flow within the website better - delivered on a roadmap plan & wireframes on the proposed solution.
​
DPC, South Australian Government
Student Technology Consultant for Digital Transformation course in CMU
Worked as a Project Manager with a POC from the department of premier & cabinet team in the South Australian Government to understand & provide solutions for a problem statement defined by the team.
Because of the way information was structured in the Cabinet Office website, the end users would typically not be able to find any information they need leading to employees having to answer questions which were readily available in the website. Additionally, the layout of the website could not be changed as it was housed within the DPC website.
Our team worked to provide a solution which solved for this problem while being mindful of the constraints & the budget. We suggested integrating Microsoft Power Virtual Agent chatbot to assimilate the information & give users direct access. We further brainstormed on the solution to integrate PVA with DPC teams internal tools.
​
Heuristic evaluation, Apple Music
Human Computer Interaction for Product Managers, CMU
Extensively used Apple Music to evaluate it against the Neilsen's Ten Usability principles.
​
Deep-dived to understand the usability principles which are followed & the ones which are violated by the application.
​
Stakeholder map & Contextual inquiry, Apple Music
Human Computer Interaction for Product Managers, CMU
Created a stakeholder
map to identify all
the primary stakeholders
involved within the
Apple Music universe.
​
Zeroed in on the power users & conducted contextual inquiry based interviews to understand the users - what do they value, what are their goals, what are their needs & what influences their decisions. Created an interview guide & ran through the interview to gain insights.
​
Came up with a problem statement to solve for according to themes noticed in the interview.
​
"To improve the community aspect of Apple Music by allowing users easy access to public playlists & nudging users to publish their own playlists"
_edited.png)
Affinity clustering & journey mapping
Human Computer Interaction for Product Managers, CMU
Used interview notes to create
affinity clusters to zero in
on the major themes/needs
the user identifies with.
​
Used these clusters to map a persona customer journey map identifying the goal, need & finally the opportunity within every step in the customer journey map.
​
Speed dating & lo-fi prototyping
Human Computer Interaction for Product Managers, CMU
Speed Dating is a design
method used to rapidly
evaluate multiple concepts
so that you can learn about how
people react to these ideas while
taking into account contextual and
social factors.
​
"As a user, I want to share the music I so lovingly made with the rock community"
​
"As a user, I want the perfect playlist for board-game night"
​
In this project, I worked on creating 3 story boards to address these user needs, with higher risk factor in each one. I then interviewed our customers by showing them the storyboards, to get quick feedback on the solutions. After that I worked on creating the lo-fi prototype of the chosen solution on Figma.
​
Think Aloud User Testing & Final Prototype
Human Computer Interaction for Product Managers, CMU
Planned the script & conducted a think-aloud interview with users. The idea was to get usability feedback on the lo-fi prototype, understand major usability pain areas, & re iterate. The users were given 3 tasks to do via the prototype.
​
The pain areas were used to improve the final prototype.
​
PearPop - Strategic deepdive
Technology Strategy, CMU
Pearpop is a creator-collaboration platform that connects users with influencers on social media to create social media content. This allows users and small businesses to gain popularity by working with a celebrity/famous content-creator and monetizing the content they create. The firm has gained a lot of traction since its inception in 2020 and now has 100000+ creators. They currently have a funding of around 17M dollars. We chose this startup for our project.
​
In this project, me & my team brainstormed on possible strategies for PearPop to grow its network effects & monetize. We also looked at various platform design improvements to supplement their next phase of growth.
​
Look at our in-depth solution here.
Building Blockchain
Distributed systems, CMU
In this project, I built a stand-alone blockchain & a client which interacted with the blockchain which has its logic running on the server. The code ran proof of work for every new block which needed to be added, once a successful hash of needed difficulty was found, it added the block to the chain.
​
There were other functionalities in the project - which verified all the blocks in the chain, added a malicious block in the chain & then repaired the blockchain.
​
Reach out to me, to have a look at the code.
Technology used: Java
EQUIP - Product Pitch
Student Product Manager for Agile Methods course in CMU
Worked with a team to ideate on a customer problem & to identify the user segment we would be solving for & developed + prioritized on a product roadmap divided in to actionable sprints. Also worked to define the organizational structure following scrum guidelines & setting up a customer service team + IVR, for the fictional product.
​
EQUIP - Gym on wheels is an application which would help deliver on gym equipment's & personal trainers at home for an easy, anxiety free exercise.
​
We believe that the concept of gym has a social pressure associated to it - of perfect bodies & pre knowing the complicated equipment's that it comes with. Working out should be an easier, peaceful experience. EQUIP works to make that happen.
​
Also finished the Scrum Product Owner certification as a part of this course.
Hangman -Twister Application
Java Project part of the OOP course in CMU
Built a Java Applet which allowed end user to play either Hangman or Twister, it picked a word from a pre defined list of words & followed the general game rules of both the games. It further stored the outcomes of all game rounds & allowed the user to have a look at his/her performance.
​
You can have a look at the code on my GitHub.
Framingham Heart Study, Data analysis
Business Intelligence, CMU
In this project, we deep dive into
The data source gives us an insight on the cause & age of death of some people along with
their other health data. If we could correctly predict what could potentially be the age of death of other people within the data set, we could help them in taking preventive measures & precautionary treatments to avoid the disease.
​
Clustering is used to find similarity patterns in data & group the data in similar clusters. We
could also use clustering to find similar health patterns among patients and add them in
clusters. Similarly, regression which is used to do prediction on a data variable, could help us
predict the potential age of death of these patients.
​
Look at our solution here.
​
​
​
BigBasket - User Retention
Product Brainstorm, Side Project
BigBasket is a grocery delivery application/product which is famous locally in India with an already established user base, as a part of this side project I worked out on 3 possible product solutions with an objective to increase user retention.
I surveyed the customers, & zeroed in on the most important problem - users considered BB as an option - they could buy potato chips from BB but they could also buy it from Amazon, Flipkart, Grofers etc. There was a lack of loyalty to commit to a single product & this is what I decided to solve for, have a look at an in-depth analysis & solution Here.
Web scraping
Distributed Systems Course, CMU
The project was mainly aimed at learning how to web scrape an HTML response to get necessary information & analyze it, to deliver something of value in our Web page.
​
In this Project, we built a website, which displayed the number of Olympic medals which were won by 20 countries & the expected number of medals by these same countries considering their GDP (G), Population(P) using the regression formula: S = 0.1*(P*G^2)^(1/3).
​
The final website displayed data scraped from 3 different websites. It showed all 20 countries along with their medals won, expected medals & a GIF of the country flag.
​
Reach out to me, to have a look at the code.
Technology used: HTML, JSP, Java, Java Servlets
TCP/UDP Client-Server architecture
Distributed Systems Course, CMU
The project was mainly aimed at learning the mechanisms of building a TCP & a UDP connection using Java Sockets.
​
In this Project, we built a client & a server using Java Sockets, and created a TCP & a UDP connection between them. We also learnt to build our own HTTP connection on top of the TCP connection. The aim of the project was to send arithmetic commands to the server using the TCP/UDP connection, process it in the server end & pass the result back to the client end.
​
Reach out to me, to have a look at the code.
Technology used: Java
Mento - User Journey identification
Product Brainstorm, Side Project
Mento is a product which has cohort based lectures on non conventional career options - think Coursera for becoming a chef or an actor.
​
As a part of this side project, I deep dive in to Mento & identify the key objectives I would work on while solving for how the User Journey would look like for a the primary customer segment starting from onboarding to a successful transaction. We also try to decipher the various emotional highpoints & lowpoints in this user journey & how product could have an direct correlation to that.
Have a look Here.
UX Comparison - Swiggy & Zomato
Product Brainstorm, Side Project
Zomato & Swiggy are both uber-popular food delivery applications used for food delivery in India. They make up for a huge portion of the complete user base in this segment.
​
As Zomato prepares for an upcoming IPO, I have a look at the key differences in User Experience / Product design between both these competing applications.
​
Have a look at the UX Comparison Here.
Cue Math - Referral Program
Product Brainstorm, Side Project
CueMath is a product which gamifies Math to help kids adapt & learn Math quickly.
​
In this product brainstorm, I worked on developing & Improving a referral program for CueMath making an assumption that the # of referrals were low & the NPS of the product was good - meaning customers liked the product but still didn't refer.
​
Have a look Here
Queue less Shopping System
BTech final Project in Java & Python, KJSCE
Worked with a team on building a prototype for a queue less shopping system. The prototype used weight sensors to detect the object being picked & an image processing algorithm to map every object with a person who picked it & added it to his/her cart. While the user wanted to leave the store, on scan the items in the cart would be automatically checked out.
​
​
Look at the report for the project Here
Offline Speech to Text Translator
Application, Mini Project, Third Year, KJSCE
Built a prototype which allowed user to enter a French word & the application automatically converts it in to English & gives a speech output. Translation + text to speech conversion is done offline
​
Have a look here at the Git for the code
