Public Feature Requirements - II
Both moderators and Automators
For the first 1000 articles, moderators should collect them After 1000 articles, only 20% work should be done by the moderators. Everything else should be taken care of our ML Engine.
It should be deliverable in Twitter, LinkedIn, Instagram
Salagar's Test
Learning-SQL.pdf
1MB
PDF
Learning SQL
PDF Viewer in posts like above
Get text from image table
PR News Analyzer:
Find Fake ML'er
Country Code to Country Converter
The existing python tool is not good enough
randomfood
I need to add More items
Create better PDF for ML posts
Tag Creator for FPR posts.
Input:
Automate Your Python Script with Process Manager 2 (PM2)
Step by Step Guide With an Example of COVID-19 Data Scraping
Output:
TBD
I have 2 months transactions of my random client. Just do a pie chart and tell me where did I spend much
Admin:
Convert as a task and get NDA
Collect specific posts like this:
Get Song likes every 2 mins by using PM2 and Python
Kaggle Ranking and Color change. Independent research
This app will collect information about your partner and will remind you often like:
What is her/his favorite food?
It can be a best tool for the first year couple or the person who forgets about his partner
RTG Questions in Typeform like UI
Job Requirement Analyzer
P1 - P5: 50, 100, 150, 250, 400 = 950
TBW
TBW
Analyze Kaggle Description:
I am Passionate about Data Science and Machine Learning and I love to work on data and data related Projects. I have specializations in Machine Learning, Data Analytics, Business Analytics, Data Visualizations. I have hands on with Data Science tools such as Jupyter Notebooks. I have done Projects under Classification, Regression, Clustering, Image Classification, Recommendation Systems, Data Analysis, Data Visualizations, Data Mining etc. I have completed my B.tech in Computer Science from ABES Engineering College, Ghaziabad.
Image processing Parallel work
I have 200 images. I need to do some custom process work parallelly in Python (with Celery) and improve the speed. 1. Make it sequence based - process images one by one 2. Try to find parallel processing and improve performance
Junk or Not
Find whether the food is Junk food or not
P1..P5: 50, 100, 150, 250, 400
Product Classifier:
Collect product description of 2 categories (Beauty, Beer). Create a binary classifier in Keras
P1..P5: 50, 100, 150, 250, 400
Images
Collect ML requirements from freelancing platforms and classify Vague or Not Vauge
Find MLM by first 3 statements
ML Post Emitter:
- Should generate #tags
- Should check spelling mistake
- Should have custom NER
- Should
P1: Just generate default tags for Featurepreneur
P2: Auto tagging
I have a zip file which contains 50 images. You need to rename the existing files with 1-100 (or any random numbers) and give me as a zip file.
Supporter or Opposer
Find this video is supporting Amway or opposing it. You should use ML to do that.
I need to get an email every day to say the market is red or not.
Changing Background by analyzing voice
Analyze keys and bring a free image in the background
Image2table
Convert Image to Table. Get the table cells in text format.
Table or Not
Classify the image whether the image is table or not
Sample table image:
Sample non-table image:
Classify weapon bundles
By using a simple classification on the weapon bundle, we should be able to classify whether the bundle is Halloween bundle or Christmas bundle
It will avoid us sending the wrong weapon bundle while publishing.
Collect various linkedin posts including tech.
Classify tech or non-tech among those
Collect some dataset from here
Collect only above 5000*3000 images in Python
Automate our existing Internship Certificate
P1: Base certificate
P2: Add hidden unique code so we can verify the certificate later
Analyze how much % false positive am I getting from Neuvoo.com for "Data Scientist"?
What is false positive
InterviewGation Improvements
Improve the interviewgation app
P Manikanta- [email protected]
Trinadh Reddy - [email protected]
"
"B Adityan, [email protected]
K Gowtham, [email protected]"
"ANGUVINAYAGAM [email protected]
Tamilselvan
Janani
Harshini
"
"1. NAVEENKUMAR S
2. SELVA SURYA C"
Surya kumar
Sadhana
Kapilan, [email protected]
Pooja
2
Create a simple email finder
IWould App:
It's a survey like app which would let you fix the future price for a product.
Right now, Netflix in my region costs 15$/month. I would subscribe if it is 8$/month
Youtube premium costs $11/month. I would subscribe if 6$/month
Social Score
We need a small app to highlight helping mindset people with social score. More social score would let them to get some discounts like student discounts.
Zamzar like work
I need a reviewer engine like this
Collect nice profile pics and analysis on them.
For Daily standup, I need to show random order with some UI and animaton
P1: Show with a good UI
P2: Identify the voice and then highlight the current speaker and move on to the next when it is done
Transcribe Daily Standup
StandupYov
Poster Analysis
Headshot Analysis
Make AI to create a video for this
Image Processing work:
Problem 1:
It takes 2 to 7 hours on the Image processing work because of ImageMagick IO issue. If we can create a better library than ImageMagick which can be integrated with Python, it would be great.
Final outcome would be a new library like ImageMagick with higher performance with Dask and Numpy integration.
tag: research
Poster Validator:
After creating a poster, we have to validate against our data to avoid mistake like these:
Madhumitha vs Madumitha
TBW
AWS Cost Analyzer
Analyze AWS Cost from the receipts
I have a receipt PDF. I need to erase the confidential information like name, adddress of the receiver by using PDF.
Turtle Score
In BDB events, we are giving 6 pushups as a challenge. We don't have a clear proof whether someone is doing a push up or not.
Kaipulla Gun
Throw various funny scnes in the background when we talk in Just Google It session.
Meme Developer Hiring
You should be good at
AI based wallpaper.
You tell soemthing. We understand your meaning and find the right images for you.
Just Google IT (JGI)
Need to integrate the tech quiz with NLP
Find the statement depth L1, L2, and L3
If L1, give low factor L2, give medium factor L3, give high factor
These decision making should be done by Human and NLP Engine
Over time, these decision making should be done by NLP Engine.
You need to collect everyone's link and then find the highest speed by using OCR and other libraries.
TBW
Transcribe Small discussion with our clients (private data)
Phases:
1. Manual Transcription
2. Automate it with AWS/Azure/Google/PyKaldi
3. Auto title creation
4. Auto tag creation
MoviesCount
Popularity
AwardsCount
YearsInTheIndustry
RoleDiversityScore
InstagramFollowers
TwitterFollowers
Collect 200 food photos
180 good photos, 20 low quality photos
Predict the quality photos
Quote maker like this
- 1.Random image picker.
- 2.Change based on the text
Ahmed Index:
We need to find similar Indian Scientists who changed the world
https://en.wikipedia.org/wiki/Nasir_Ahmed_(engineer)
CarryYou Index:
Theranos skeptist similar people
I need page like this to show KAPI (Kollywood API)
I need to automate the cerficate making in BDB, Intern via Herokuapp. Create a Flask app and automate this work.
Find the similar singers to SPB
Find the similar Music Directors to Ilaiyaraaja
Find the similar Directors to Mani Ratnam
Collect lyrics from videos like this by using Tensorflow
Find similar players of MS
WordBuilding with NLP Burst:
People say words
Our engine will listen to them and show the word with bursting cloud (like crackers) with points related to the words
How to get points for each word?
That's where Tact Scoring Engine comes as Unique Selling Element
Created by: Lil Boys
Spelling mistakes in Trump tweets
Followers, Retweet counts of Trump
I have some recorded videos which might contain AWS sensitive information. You have to write a code in Tensorflow to find whether I do have AWS sensitive information or not.
If sensitive information is available, show the section of in the video.
I have a video that should be uploaded on YouTube. When I try to publish, youtube gives me only limited thumbnails which are not enough for me.
You shoud write a code to emit 20-50 thumbnails for my video so I can post the best thumbnail on YouTube.
We want to have a symbol for our Tact Coins. You can create some currency symbols like below and share with us.
By using any image proessing technique or ML algorithm/lib, get each state with population.
Collect best lines from TV shows like these:
Two and Half Men
How I Met Your Mother
Brooklyn Nine-Nine
The Modern Family
Do some sentiment analysis on those best lines and put them in the table.
P1: 25 Lines
P2: 50 Lines
Vivekanandha Index
Collect various thinkers like Vivekanandha with simple variables.
We have to find which team member has the high speed internet. For this, we ask everyone to send a link of their speed test result like below:
ImageMagick
Do some sample work with ImageMagick.
SICA Effect and Research
We need some volunteers to work on our SICA effect and research work. We will explain things in person.
Ontario addresses like in this format N0N 0N0. We need to come up with this format in South Indian states and addresss.
Come up with algorithm to change all pincodes to new format.
Requirements for JGI (Just Google It)
1.Login for each team
2.Enter Question(by each team)
3.Display question ( Question to be displayed after certain time limits 2 mins - answering, the next qn displaying)
4.Storing JGI topic and Question (Can be used for future use) 5.Scoring for admin
Tact Among Us
1.Choosing random words
2.Choosing crew mate and imposter
3.Decide number of rounds
4.Order of participants
5.Emergency meeting for each round/each person
Requirements for tulkot synge :
1.Questions are displayed in the screen .
2.There are certain time limit for answering .
3.First,they want to find the song .
4.The answering team has to sung the song (if expectation matches extra points awarded) . 5.Storing songs and sentences (Can be used for future use).
Tact-Cricket:
Max score : Winner
Scores range : [0,2,4,6,1,3]
Out : 0 (appearing on the screen)
Score : Cummulative adding of the numbers
Achievements :
Continuous 3 fours = +6
Continuous 3 sixes = +9
Continuous 6 fours = +12
Continuous 6 sixes = +18
Thooki Jail la podunga sir (Game)
Requirements
1. each member has to tell a letter to find out the movie/technical
2. If letter told by the person is crct - Fill those letter in empty spaces
3. If wrong for the first time face of the man will be drawn, then hand, then leg, then put in jail If they didnt find the answer within that chance,out of the game
AnyTask with Celery
Image processing, Video conversion, Video anaysis, take any task. It should be done by our app with external plugins where developers can add plugins and inject them.
Status: In Progress
Feature Holder: Team Celery
Find similar nobel prize winners of Solow and make a visual about them.
Gitbook table to REST API:
I need to convert simple table to API.
This would be used for spin the wheel winners. So, I can store the details in the Gitbook and then give a link to Front end guys to get the work done ASAP.
TBD
TBD
TBD
For more see: Public Feature Requirements III
Last modified 2yr ago