Niranjan Lamichhane
@niranjannlc.bsky.social
Learning Artificial intelligence.
Enrolled in #100DaysOfCode
Enrolled in #100DaysOfCode
day 71 of #100DaysOfCode
Used hex platform to analyze the migration platform of nepla .
have thought of making final submiison on the hackathon .
Used hex platform to analyze the migration platform of nepla .
have thought of making final submiison on the hackathon .
January 21, 2026 at 2:40 PM
day 71 of #100DaysOfCode
Used hex platform to analyze the migration platform of nepla .
have thought of making final submiison on the hackathon .
Used hex platform to analyze the migration platform of nepla .
have thought of making final submiison on the hackathon .
Day 70 of #100DaysOfCode
Getting comfortable with hex platform to do ai driven data analytics.
Learning to construct stories from data .
Getting comfortable with hex platform to do ai driven data analytics.
Learning to construct stories from data .
January 20, 2026 at 5:36 AM
Day 70 of #100DaysOfCode
Getting comfortable with hex platform to do ai driven data analytics.
Learning to construct stories from data .
Getting comfortable with hex platform to do ai driven data analytics.
Learning to construct stories from data .
day 69 of #100DaysOfCode
Contuing working with a tip prediction model with 7 features, then discovered 5 had severe multicollinearity (VIF > 10) 😱
Pruned to 5 independent features. Result: clean coefficients + competitive accuracy.
Contuing working with a tip prediction model with 7 features, then discovered 5 had severe multicollinearity (VIF > 10) 😱
Pruned to 5 independent features. Result: clean coefficients + competitive accuracy.
January 18, 2026 at 7:04 AM
day 69 of #100DaysOfCode
Contuing working with a tip prediction model with 7 features, then discovered 5 had severe multicollinearity (VIF > 10) 😱
Pruned to 5 independent features. Result: clean coefficients + competitive accuracy.
Contuing working with a tip prediction model with 7 features, then discovered 5 had severe multicollinearity (VIF > 10) 😱
Pruned to 5 independent features. Result: clean coefficients + competitive accuracy.
day 68 of #100DaysOFCode
Just following the quote , " Sharpen your axe before cutting the tree. "
in abstract term , just sharpening the axe today for hackathon project .
Just following the quote , " Sharpen your axe before cutting the tree. "
in abstract term , just sharpening the axe today for hackathon project .
January 17, 2026 at 1:06 PM
day 68 of #100DaysOFCode
Just following the quote , " Sharpen your axe before cutting the tree. "
in abstract term , just sharpening the axe today for hackathon project .
Just following the quote , " Sharpen your axe before cutting the tree. "
in abstract term , just sharpening the axe today for hackathon project .
DAY 67 OF #100DaysOfCode
Didnot write a code , just learning to tell the story from the data in the form of contradictions, accelerations , turning point and underdog .
Didnot write a code , just learning to tell the story from the data in the form of contradictions, accelerations , turning point and underdog .
January 16, 2026 at 11:52 AM
DAY 67 OF #100DaysOfCode
Didnot write a code , just learning to tell the story from the data in the form of contradictions, accelerations , turning point and underdog .
Didnot write a code , just learning to tell the story from the data in the form of contradictions, accelerations , turning point and underdog .
January 15, 2026 at 2:07 PM
day 65 of #100DaysOfCode
Computing and interpreting R² for the simple linear model.
Building and reading the actual vs predicted plot.
Building and interpreting the residual plots and diagnostics, and concluding that more features are needed beyond total_bill_log in my model .
Computing and interpreting R² for the simple linear model.
Building and reading the actual vs predicted plot.
Building and interpreting the residual plots and diagnostics, and concluding that more features are needed beyond total_bill_log in my model .
January 14, 2026 at 1:23 PM
day 65 of #100DaysOfCode
Computing and interpreting R² for the simple linear model.
Building and reading the actual vs predicted plot.
Building and interpreting the residual plots and diagnostics, and concluding that more features are needed beyond total_bill_log in my model .
Computing and interpreting R² for the simple linear model.
Building and reading the actual vs predicted plot.
Building and interpreting the residual plots and diagnostics, and concluding that more features are needed beyond total_bill_log in my model .
day 64 of #100DaysOfTarget
Constructing the simple model with the one feature and comparing its with existing baseline model of mean and 15% heuristic .
Result is : The simple model outperforms both baselines, improving MAE by ~9.5% and surpassing the 5% complexity threshold.
Constructing the simple model with the one feature and comparing its with existing baseline model of mean and 15% heuristic .
Result is : The simple model outperforms both baselines, improving MAE by ~9.5% and surpassing the 5% complexity threshold.
January 13, 2026 at 7:24 AM
day 64 of #100DaysOfTarget
Constructing the simple model with the one feature and comparing its with existing baseline model of mean and 15% heuristic .
Result is : The simple model outperforms both baselines, improving MAE by ~9.5% and surpassing the 5% complexity threshold.
Constructing the simple model with the one feature and comparing its with existing baseline model of mean and 15% heuristic .
Result is : The simple model outperforms both baselines, improving MAE by ~9.5% and surpassing the 5% complexity threshold.
day 63 of #100DaysOfCode
Added 3 engineered features (bill_per_person, is_weekend, time_dinner) with statistical validation. Captures financial, temporal & experiential context for better predictions.
Added 3 engineered features (bill_per_person, is_weekend, time_dinner) with statistical validation. Captures financial, temporal & experiential context for better predictions.
January 11, 2026 at 1:27 PM
day 63 of #100DaysOfCode
Added 3 engineered features (bill_per_person, is_weekend, time_dinner) with statistical validation. Captures financial, temporal & experiential context for better predictions.
Added 3 engineered features (bill_per_person, is_weekend, time_dinner) with statistical validation. Captures financial, temporal & experiential context for better predictions.
day 62 of #100DaysOfCode
Learned that Skewness vs Mean-Median diff : High skewness can exist despite close mean/median (extreme outliers)
Reduced skewness from 1.12 to -0.12 (nearly normal) by using log transformation
Applied Encoding strategy dependent on feature type - Binary vs one-hot
Learned that Skewness vs Mean-Median diff : High skewness can exist despite close mean/median (extreme outliers)
Reduced skewness from 1.12 to -0.12 (nearly normal) by using log transformation
Applied Encoding strategy dependent on feature type - Binary vs one-hot
January 10, 2026 at 3:02 PM
day 62 of #100DaysOfCode
Learned that Skewness vs Mean-Median diff : High skewness can exist despite close mean/median (extreme outliers)
Reduced skewness from 1.12 to -0.12 (nearly normal) by using log transformation
Applied Encoding strategy dependent on feature type - Binary vs one-hot
Learned that Skewness vs Mean-Median diff : High skewness can exist despite close mean/median (extreme outliers)
Reduced skewness from 1.12 to -0.12 (nearly normal) by using log transformation
Applied Encoding strategy dependent on feature type - Binary vs one-hot
day 61 of #100DaysOfCode
Set a 5% relative improvement rule to justify model complexity. Built is_worth_complexity() to compare all future models against the 15% Rule oracle (MAE 4.0061%).
Playing with visualization to analyze and express variance
Set a 5% relative improvement rule to justify model complexity. Built is_worth_complexity() to compare all future models against the 15% Rule oracle (MAE 4.0061%).
Playing with visualization to analyze and express variance
January 9, 2026 at 12:22 PM
day 61 of #100DaysOfCode
Set a 5% relative improvement rule to justify model complexity. Built is_worth_complexity() to compare all future models against the 15% Rule oracle (MAE 4.0061%).
Playing with visualization to analyze and express variance
Set a 5% relative improvement rule to justify model complexity. Built is_worth_complexity() to compare all future models against the 15% Rule oracle (MAE 4.0061%).
Playing with visualization to analyze and express variance
day 59 of #100DaysOfCode
Not any post worthy progress or achivement .
just learned to condense the work being done to documentable achivement.
Not any post worthy progress or achivement .
just learned to condense the work being done to documentable achivement.
January 7, 2026 at 12:33 PM
day 59 of #100DaysOfCode
Not any post worthy progress or achivement .
just learned to condense the work being done to documentable achivement.
Not any post worthy progress or achivement .
just learned to condense the work being done to documentable achivement.
Day 58 of #100DaysOfCode
Yeah it's working.
From tomorrow onwards, I will be posting my updates through this way only .
Yeah it's working.
From tomorrow onwards, I will be posting my updates through this way only .
Hi i am testing from bluesky api posting . is it really working
January 6, 2026 at 3:49 PM
Day 58 of #100DaysOfCode
Yeah it's working.
From tomorrow onwards, I will be posting my updates through this way only .
Yeah it's working.
From tomorrow onwards, I will be posting my updates through this way only .
Hi i am testing from bluesky api posting . is it really working
January 6, 2026 at 3:47 PM
Hi i am testing from bluesky api posting . is it really working
Day 57 of #100DaysOfCode
Going through some theory only today...
I would no longer wonder what the hell hyperparameter would be.
Going through some theory only today...
I would no longer wonder what the hell hyperparameter would be.
January 5, 2026 at 2:04 PM
Day 57 of #100DaysOfCode
Going through some theory only today...
I would no longer wonder what the hell hyperparameter would be.
Going through some theory only today...
I would no longer wonder what the hell hyperparameter would be.
Day 56 of #100DaysOfCode
Just brushing up and making basic concepts of machine learning so that it contribute for stronger foundations.
Constructed the Duolingo like app to test my knowledge on those concepts.
Just brushing up and making basic concepts of machine learning so that it contribute for stronger foundations.
Constructed the Duolingo like app to test my knowledge on those concepts.
January 1, 2026 at 12:24 PM
Day 56 of #100DaysOfCode
Just brushing up and making basic concepts of machine learning so that it contribute for stronger foundations.
Constructed the Duolingo like app to test my knowledge on those concepts.
Just brushing up and making basic concepts of machine learning so that it contribute for stronger foundations.
Constructed the Duolingo like app to test my knowledge on those concepts.
Day 55 of #100DaysOfCode
You don't win all the time .
Bugs failure are part of journey.
So today is the one for me .
Failed to integrate datadog and submit for hackthon due to quota exhaustion error from Google gemini.
😭😭
You don't win all the time .
Bugs failure are part of journey.
So today is the one for me .
Failed to integrate datadog and submit for hackthon due to quota exhaustion error from Google gemini.
😭😭
December 31, 2025 at 12:08 PM
Day 55 of #100DaysOfCode
You don't win all the time .
Bugs failure are part of journey.
So today is the one for me .
Failed to integrate datadog and submit for hackthon due to quota exhaustion error from Google gemini.
😭😭
You don't win all the time .
Bugs failure are part of journey.
So today is the one for me .
Failed to integrate datadog and submit for hackthon due to quota exhaustion error from Google gemini.
😭😭
Day 53 and 54 of #100DaysOfCode
Ideating for implementing datadog and participating in datadog challenge.
Ideating for implementing datadog and participating in datadog challenge.
December 30, 2025 at 10:49 AM
Day 53 and 54 of #100DaysOfCode
Ideating for implementing datadog and participating in datadog challenge.
Ideating for implementing datadog and participating in datadog challenge.
Reposted by Niranjan Lamichhane
"Solo founder" sounds romantic until you realize it means:
- No one to cover when you're sick
- Every decision paralyzes you
- Celebrating wins alone
- Burnout has no backup
The indie hacker myth: you're self-sufficient.
The reality: you just can't afford help yet.
- No one to cover when you're sick
- Every decision paralyzes you
- Celebrating wins alone
- Burnout has no backup
The indie hacker myth: you're self-sufficient.
The reality: you just can't afford help yet.
December 28, 2025 at 7:00 PM
"Solo founder" sounds romantic until you realize it means:
- No one to cover when you're sick
- Every decision paralyzes you
- Celebrating wins alone
- Burnout has no backup
The indie hacker myth: you're self-sufficient.
The reality: you just can't afford help yet.
- No one to cover when you're sick
- Every decision paralyzes you
- Celebrating wins alone
- Burnout has no backup
The indie hacker myth: you're self-sufficient.
The reality: you just can't afford help yet.
Day 52 of #100DaysOfCode
Experimented with correlations matrix and pairplot.
Updated the blog about my learning journey of data science by experimenting with tips dataset.
Blog link : dev.to/niranjannlc/...
Experimented with correlations matrix and pairplot.
Updated the blog about my learning journey of data science by experimenting with tips dataset.
Blog link : dev.to/niranjannlc/...
My Data Science Journey: Restaurant Tips Analysis
Project: Exploratory Data Analysis on Restaurant Tips Dataset Duration: Full EDA Process Dataset: 243...
dev.to
December 28, 2025 at 11:14 AM
Day 52 of #100DaysOfCode
Experimented with correlations matrix and pairplot.
Updated the blog about my learning journey of data science by experimenting with tips dataset.
Blog link : dev.to/niranjannlc/...
Experimented with correlations matrix and pairplot.
Updated the blog about my learning journey of data science by experimenting with tips dataset.
Blog link : dev.to/niranjannlc/...
Day 51 of #100DaysOfCode
Just working on testing hypothesis, using visualization tool, gathering insight by interpretation on tips datasets.
Will work on correlation matrix , heatmap and pairplot tomorrow.
Just working on testing hypothesis, using visualization tool, gathering insight by interpretation on tips datasets.
Will work on correlation matrix , heatmap and pairplot tomorrow.
December 27, 2025 at 11:48 AM
Day 51 of #100DaysOfCode
Just working on testing hypothesis, using visualization tool, gathering insight by interpretation on tips datasets.
Will work on correlation matrix , heatmap and pairplot tomorrow.
Just working on testing hypothesis, using visualization tool, gathering insight by interpretation on tips datasets.
Will work on correlation matrix , heatmap and pairplot tomorrow.
Day 50 of #100DaysOfCode
-Trying to construct own framework to turn hypotheses into visualization strategies in datasets .
-Practiced choosing charts (scatter, box, bar) by variable type and goal, and used boxplots (IQR, whiskers, outliers) to analyze skewness, symmetry, and relationships in data.
-Trying to construct own framework to turn hypotheses into visualization strategies in datasets .
-Practiced choosing charts (scatter, box, bar) by variable type and goal, and used boxplots (IQR, whiskers, outliers) to analyze skewness, symmetry, and relationships in data.
December 26, 2025 at 11:03 AM
Day 50 of #100DaysOfCode
-Trying to construct own framework to turn hypotheses into visualization strategies in datasets .
-Practiced choosing charts (scatter, box, bar) by variable type and goal, and used boxplots (IQR, whiskers, outliers) to analyze skewness, symmetry, and relationships in data.
-Trying to construct own framework to turn hypotheses into visualization strategies in datasets .
-Practiced choosing charts (scatter, box, bar) by variable type and goal, and used boxplots (IQR, whiskers, outliers) to analyze skewness, symmetry, and relationships in data.
Day 49 of #100DaysOfCode
Developed gamified app to track the todo list using Google AI studio
Learned to observe the skewness of data of dataset and dealing with missing values in datasets.
Developed gamified app to track the todo list using Google AI studio
Learned to observe the skewness of data of dataset and dealing with missing values in datasets.
December 24, 2025 at 10:49 AM
Day 49 of #100DaysOfCode
Developed gamified app to track the todo list using Google AI studio
Learned to observe the skewness of data of dataset and dealing with missing values in datasets.
Developed gamified app to track the todo list using Google AI studio
Learned to observe the skewness of data of dataset and dealing with missing values in datasets.
Day 48 of #100DaysOfCode
Build and deployed my project " Dastabsej to site " as a warmup task for ernie ai developer challenges .
Check my GitHub repo : github.com/NiranjanNlc/...
Build and deployed my project " Dastabsej to site " as a warmup task for ernie ai developer challenges .
Check my GitHub repo : github.com/NiranjanNlc/...
December 23, 2025 at 12:13 PM
Day 48 of #100DaysOfCode
Build and deployed my project " Dastabsej to site " as a warmup task for ernie ai developer challenges .
Check my GitHub repo : github.com/NiranjanNlc/...
Build and deployed my project " Dastabsej to site " as a warmup task for ernie ai developer challenges .
Check my GitHub repo : github.com/NiranjanNlc/...