TIKVAH-SPORT
α¨α¨αα³ α α°ααα ! 4:00 αα¨αα³α α¨ α΄ααα @tikvahethsport @kidusyoftahe
The second half started π΄π΄π΄
It is hard to guess who will win
Good night famπ«‘π«‘
It is hard to guess who will win
Good night famπ«‘π«‘
π«‘πNew challenge this week!
Building a Retrieval Augmented Generation (RAG) system that can answer questions from customer complaint data using semantic search and LLMs.
The goal is to transform thousands of customer complaints into meaningful insights that help teams better understand customer pain points.
This week project will lead me to learn more about vector databases, embeddings, retrieval systems, and AI-powered question answering.
πDAY 1
#AI #RAG #MachineLearning #LLM #GenerativeAI
Building a Retrieval Augmented Generation (RAG) system that can answer questions from customer complaint data using semantic search and LLMs.
The goal is to transform thousands of customer complaints into meaningful insights that help teams better understand customer pain points.
This week project will lead me to learn more about vector databases, embeddings, retrieval systems, and AI-powered question answering.
πDAY 1
#AI #RAG #MachineLearning #LLM #GenerativeAI
β€2
Here is the real Leoππ½βπΏ
https://t.me/leotriesdesign/787
https://t.me/leotriesdesign/787
Telegram
Leo Design (Non - Design)
@leotriesdesign
β€3
As always well said π€πΎ and noted !βπΏ
https://t.me/selfmadecoder/4154
https://t.me/selfmadecoder/4154
Telegram
Tech Nerd
Most people wonβt see the vision in the beginning. not because theyβre against you, but because they can only judge what already exists.
They see your current skills, your current results, and your current circumstances. From that, they form an opinion ofβ¦
They see your current skills, your current results, and your current circumstances. From that, they form an opinion ofβ¦
β€2
Simret.exe
π«‘πNew challenge this week! Building a Retrieval Augmented Generation (RAG) system that can answer questions from customer complaint data using semantic search and LLMs. The goal is to transform thousands of customer complaints into meaningful insights thatβ¦
This phase needs a deep understanding of the setup you really have to know which packages to install and how to balance dependencies if you don't want to cry later! π
The next step is running EDA on a massive customer complaint dataset across credit cards, personal loans, and money transfers. I hope it isnβt going to be too hard! π€βοΈ
Good night everyone! π΄β¨
The next step is running EDA on a massive customer complaint dataset across credit cards, personal loans, and money transfers. I hope it isnβt going to be too hard! π€βοΈ
Good night everyone! π΄β¨
β€2
β¨Today's New concept
Approximate Nearest Neighbor (ANN) algorithms make similarity search much faster by finding vectors that are close enough to the nearest neighbors without checking every vector in the database.
#DAY2 of the project
Approximate Nearest Neighbor (ANN) algorithms make similarity search much faster by finding vectors that are close enough to the nearest neighbors without checking every vector in the database.
#DAY2 of the project
π₯3
Word of the day π«°
https://t.me/abroid0/408
https://t.me/abroid0/408
Telegram
Abroid ΓΔv
one of my favorite movie scenes of all time is from "The Pursuit of Happyness".
during an interview, he says:
i'm the type of person, if you ask me a question and i don't know the answer, i'm gonna tell you that i don't know. but i bet you what: i knowβ¦
during an interview, he says:
i'm the type of person, if you ask me a question and i don't know the answer, i'm gonna tell you that i don't know. but i bet you what: i knowβ¦
Forwarded from Yohannes Haile (Yohannes Haile)
βThe undocumented life is not worth livingβ
β Socrates, Influencer
β Socrates, Influencer
π2
This shows mom's resilience has no limits.β€οΈβ€οΈ still she is young but she achieve her dreams
β€5
π«‘Project update on #DAY2
I have been working on :
Loading & explored 9.6 million real CFPB customer complaintsπ€§π€§π€§
Analyzed complaint distribution across 21 product categories
Built a text cleaning pipeline (boilerplate removal, normalization)
Set up CI/CD with GitHub Actions β 7 tests passing β
Next up: building the embedding pipeline to turn these complaints into searchable AI vectors! π₯
It seems easy but it is not.It needs careful consideration of the data.
I have been working on :
Loading & explored 9.6 million real CFPB customer complaintsπ€§π€§π€§
Analyzed complaint distribution across 21 product categories
Built a text cleaning pipeline (boilerplate removal, normalization)
Set up CI/CD with GitHub Actions β 7 tests passing β
Next up: building the embedding pipeline to turn these complaints into searchable AI vectors! π₯
It seems easy but it is not.It needs careful consideration of the data.
β€4
Good night famsπ€π«‘
Don't forget to restβΊοΈπ
Let us say this word of the night βΊοΈβΊοΈπ
https://t.me/Luna_moonmam/2793
Don't forget to restβΊοΈπ
Let us say this word of the night βΊοΈβΊοΈπ
https://t.me/Luna_moonmam/2793
Telegram
Luna's pathwayπ€
For the people who struggle alone.
For the people who carry responsibilities nobody sees.
For the people who take the blame to protect others.
For the people who stay silent while their efforts go unnoticed.
For the people who hold teams together whileβ¦
For the people who carry responsibilities nobody sees.
For the people who take the blame to protect others.
For the people who stay silent while their efforts go unnoticed.
For the people who hold teams together whileβ¦
β€1
Simret.exe
π«‘Project update on #DAY2 I have been working on : Loading & explored 9.6 million real CFPB customer complaintsπ€§π€§π€§ Analyzed complaint distribution across 21 product categories Built a text cleaning pipeline (boilerplate removal, normalization) Set up CI/CDβ¦
#DAY3 of the project
Yesterday what I did was that cleaning and organising the complaint letters. So the next part is teaching the AI how to read and remember them.It is going to be implementing text chunking , choosing an embedding model and also generating vector embedding for each chunk.
βοΈβοΈtrust the process
Yesterday what I did was that cleaning and organising the complaint letters. So the next part is teaching the AI how to read and remember them.It is going to be implementing text chunking , choosing an embedding model and also generating vector embedding for each chunk.
βοΈβοΈtrust the process
β‘3
This media is not supported in your browser
VIEW IN TELEGRAM
I have no boundaries how much I love pheven_teklayβ€οΈππ her voice her thoughts βΊοΈπ€
This one I found on my gallery like a remainder βΊοΈ
People will always have opinions and labels, but your peace is your responsibility. Guard your energy like a soldier, heal your soul like a doctor, and keep moving forward. πΆββοΈπ«
This one I found on my gallery like a remainder βΊοΈ
People will always have opinions and labels, but your peace is your responsibility. Guard your energy like a soldier, heal your soul like a doctor, and keep moving forward. πΆββοΈπ«
β€3
Forwarded from Opportunity Alertsπ’
#Opportunity_Alertsπ£
πPaid Internship Opportunity: Join Mastercard Foundation Associates Program 2026 for Young Africansπ
β¨Are you a recent graduate looking for paid work experience, mentorship, and career development opportunities? The Mastercard Foundation Associates Program is offering a 12-month paid internship for young people across East Africa.
What You'll Gain:
πΉHands-on professional experience
πΉMentorship and career support
πΉLeadership and employability skills development
πΉProfessional networking opportunities
πΉPathways to employment or entrepreneurship
Who Can Apply?
πΈRecent graduates with 0β2 years of experience
πΈMastercard Foundation Scholars Program alumni
πΈAlumni of Mastercard Foundation programs
πΈYoung people aged 18β35
πΈApplicants from Ethiopia, Kenya, Uganda, Rwanda, Tanzania, South Sudan, & Burundi
π Durations: 12 Months
πApply: https://bit.ly/sclrs26
"If this isn't for you, please share it with others who might be interested."π
Follow usπfor more opportunities
@opportunity_alerts
πPaid Internship Opportunity: Join Mastercard Foundation Associates Program 2026 for Young Africansπ
β¨Are you a recent graduate looking for paid work experience, mentorship, and career development opportunities? The Mastercard Foundation Associates Program is offering a 12-month paid internship for young people across East Africa.
What You'll Gain:
πΉHands-on professional experience
πΉMentorship and career support
πΉLeadership and employability skills development
πΉProfessional networking opportunities
πΉPathways to employment or entrepreneurship
Who Can Apply?
πΈRecent graduates with 0β2 years of experience
πΈMastercard Foundation Scholars Program alumni
πΈAlumni of Mastercard Foundation programs
πΈYoung people aged 18β35
πΈApplicants from Ethiopia, Kenya, Uganda, Rwanda, Tanzania, South Sudan, & Burundi
π Durations: 12 Months
πApply: https://bit.ly/sclrs26
Follow usπfor more opportunities
@opportunity_alerts
β€1
Simret.exe
#DAY3 of the project Yesterday what I did was that cleaning and organising the complaint letters. So the next part is teaching the AI how to read and remember them.It is going to be implementing text chunking , choosing an embedding model and also generatingβ¦
Todayβs work takes a lot of my energy π€§π€§π€§
I taught AI to understand complaints not just read them.I took 12,000 real customer complaints and converted them into numbers using a model called "all-MiniLM-L6-v2." Each complaint chunk becomes 384 numbers that capture its meaning. Similar complaints produce similar numbers so the AI can find relevant ones even when the exact words don't match.
π Next: step connecting this to a language model so it can actually answer questions like "Why are customers unhappy with Credit Cards?" using real complaint evidence.
#RAG #projectupdate #DAY3
Good night π΄ π₯± fams
I taught AI to understand complaints not just read them.I took 12,000 real customer complaints and converted them into numbers using a model called "all-MiniLM-L6-v2." Each complaint chunk becomes 384 numbers that capture its meaning. Similar complaints produce similar numbers so the AI can find relevant ones even when the exact words don't match.
π Next: step connecting this to a language model so it can actually answer questions like "Why are customers unhappy with Credit Cards?" using real complaint evidence.
#RAG #projectupdate #DAY3
Good night π΄ π₯± fams
π₯4π1