jacobjerryarackal

AI-Native Full-Stack Engineer | MERN + GenAI (RAG & Vector DBs) | Building Production-Ready AI Applications

Joined on Dec 3, 2020 https://jacobjerryarackal.vercel.app/

jacobjerryarackal

Jun 18

Transformers & The Attention Mechanism: How AI Learned to Focus

#ai #programming #python #machinelearning

6 min read

jacobjerryarackal

Jun 17

Why Most AI Agents Fail in Production And the Architecture Patterns That Actually Work

#ai #programming #datascience #machinelearning

5 min read

jacobjerryarackal

May 18

PCA (Principal Component Analysis): Finding the Hidden Structure in High‑Dimensional Data

#ai #machinelearning #datascience #programming

5 min read

jacobjerryarackal

May 15

XGBoost: When Gradient Boosting Meets Regularization

#machinelearning #ai #deeplearning #datascience

5 min read

jacobjerryarackal

May 14

Why “Just Prompting” Fails on Private Data: A RAG Post‑Mortem

#rag #ai #machinelearning #architecture

5 min read

jacobjerryarackal

Apr 24

The Copy‑Paste Trap: When Knowing “Everything” Still Leaves You with Nothing

4 min read

jacobjerryarackal

Apr 8

I Built a RAG Pipeline. Then I Realized Retrieval Is the Real Model

#ai #llm #machinelearning #rag

3 min read

jacobjerryarackal

Apr 4

We Let an LLM Control a File System and Run Commands – Here’s What Actually Broke First

3 min read

jacobjerryarackal

Apr 2

“Prompt Engineering Is Enough” Is Wrong – Here’s What I Had to Add

#ai #machinelearning #architecture #learning

2 min read

DEV Community

jacobjerryarackal

Badges

Writing Debut

Five Year Club

Four Year Club

Three Year Club

Two Year Club

One Year Club

Currently learning

Transformers & The Attention Mechanism: How AI Learned to Focus

Why Most AI Agents Fail in Production And the Architecture Patterns That Actually Work

PCA (Principal Component Analysis): Finding the Hidden Structure in High‑Dimensional Data

XGBoost: When Gradient Boosting Meets Regularization

Why “Just Prompting” Fails on Private Data: A RAG Post‑Mortem

The Copy‑Paste Trap: When Knowing “Everything” Still Leaves You with Nothing

I Built a RAG Pipeline. Then I Realized Retrieval Is the Real Model

We Let an LLM Control a File System and Run Commands – Here’s What Actually Broke First

“Prompt Engineering Is Enough” Is Wrong – Here’s What I Had to Add