Yugansh Agarwal

"Passionate coder exploring the tech world 💻 | Learning programming, solving problems, and building projects 🚀| Sharing my journey to grow with the community"

[Yugansh Agarwal @yuganshaga_dev]

Chatbot Backend API is Live 🤖🌐

Exposed the RAG pipeline via a FastAPI `/chatbot/` route.
Tested payload responses returning grounded answers with source citations.

Next step: connecting the API to the frontend client 💻

#FastAPI #RAG #WebDev #Python

[Yugansh Agarwal @yuganshaga_dev]

Caught a subtle bug in my RAG config 🐛💻

Tracked a type issue in Pydantic: EMBEDDING_DIMENSION was `str` instead of `int`.
Fixed the schema so env values parse cleanly.

Small type mismatches can quietly break entire vector pipelines ⚙️

#Python #RAG #Coding

[Yugansh Agarwal @yuganshaga_dev]

@jessicagoh2601 @piyushgarg_dev Not yet, but I'll try that too!

[Yugansh Agarwal @yuganshaga_dev]

Built a Voice-to-Voice AI Agent 🤖🎙️

Covered STS, chained architectures & TTS.
Implemented an agent that takes voice input, executes tasks & replies in voice.

Understanding latency tradeoffs across the STT-LLM-TTS pipeline 🧠

@piyushgarg_dev
#AI #SpeechAI #Python

[Yugansh Agarwal @yuganshaga_dev]

Completed Full Stack GenAI & Agentic AI with Python 🎯🤖

Covered RAG, LangGraph, agent memory & MCP.
Built hands-on agents and bridged theory into code.

Grateful to @Hiteshdotcom & @piyushgarg_dev for the guidance! Now focused on building 💻

#GenerativeAI #AgenticAI #Python

[Yugansh Agarwal @yuganshaga_dev]

Dived into Model Context Protocol (MCP) 🤖🔌

Covered MCP client-server architecture & schemas.
Explored connecting LLMs to local tools via standardized servers.

Grasping how universal protocols replace custom API glue code 🧠

Credits: @piyushgarg_dev
#AI #MCP #Coding

[Yugansh Agarwal @yuganshaga_dev]

@piyushgarg_dev Thanks Sir

[Yugansh Agarwal @yuganshaga_dev]

Explored Graph Memory using Neo4j 🤖🗄️

Covered knowledge graphs & entity relations.
Debugged query & setup errors to get agent relation-mapping working.

Grasping how graph memory captures context beyond vector similarity 🧠

@piyushgarg_dev
#AI #Neo4j #GraphDB

[Yugansh Agarwal @yuganshaga_dev]

Explored Memory Systems in AI Agents 🤖🧠

Covered short-term, factual, episodic & semantic memory.
Implemented hands-on persistence using Mem0 and Qdrant DB.

Understanding how vector storage grounds long-term recall ⚙️

@piyushgarg_dev
#AI #Qdrant #Mem0 #Python

[Yugansh Agarwal @yuganshaga_dev]

Explored LangGraph Checkpointing 🤖⚙️

Covered thread IDs & state persistence across restarts.
Built a bot to practice saving and resuming context after shutdown.

Understanding how checkpointers enable fault-tolerant agents 🧠

@piyushgarg_dev
#LangGraph #AI #Python

[Yugansh Agarwal @yuganshaga_dev]

Built a Custom Bot using LangGraph 🤖💻

Covered state graphs, nodes, edges & conditional routing.
Created a bot for hands-on practice with agent state management.

Grasping how cyclical graphs provide precise loop control 🧠

@piyushgarg_dev
#AI #LangGraph #Python

[Yugansh Agarwal @yuganshaga_dev]

Built a Multimodal Vision Agent 🤖💻

Covered vision-language models & image tokenization.
Created an agent taking image inputs to generate descriptive captions.
Understanding how visual and text tokens align in a shared embedding space 🧠
@piyushgarg_dev
#AI #ComputerVision #LLM

[Yugansh Agarwal @yuganshaga_dev]

Built a RAG Agent with FastAPI 🤖🌐

Covered chunking, vector search & context injection.
Created an agent querying custom data, served via FastAPI endpoints.

Understanding how retrieval anchors LLMs and reduces hallucination 🧠

@piyushgarg_dev
#RAG #FastAPI #AI #Python

[Yugansh Agarwal @yuganshaga_dev]

Built a Custom AI Agent in Python 🤖💻

Covered agentic workflows, tool use & reasoning loops.
Implemented an agent using Gemini LLM for multi-step task execution.

Understanding how structured workflows move LLMs beyond single prompts 🧠

@piyushgarg_dev
#AI #Python #Gemini

[Yugansh Agarwal @yuganshaga_dev]

Studied Transformer Architecture 🤖⚙️

Explored "Attention Is All You Need": tokenization, embeddings, positional encoding & multi-head attention.

Understanding how self-attention eliminates recurrence in sequence modeling 🧠

@piyushgarg_dev
#MachineLearning #AI #NLP

[Yugansh Agarwal @yuganshaga_dev]

Studied Backend Auth & Tokens 🌐⚙️

Covered JWTs, access/refresh tokens & data vs. non-data tokens.
Tested authentication vs. authorization logic and token validation.

Building intuition for secure session design 🧠

@ChaiCodeHQ @Hiteshdotcom
#Backend #WebDevelopment #Auth #JWT

[Yugansh Agarwal @yuganshaga_dev]

Learned K-Means Clustering 📊🤖
Covered the theory and implementation, including WCSS, Elbow Method & Silhouette Score for selecting the optimal number of clusters.
Building stronger intuition for unsupervised learning ⚙️
@Krishnaik06
#MachineLearning #KMeans #Clustering #Pythonv

[Yugansh Agarwal @yuganshaga_dev]

Learned Principal Component Analysis (PCA) 📉🧠
Explored dimensionality reduction, variance preservation, and feature transformation for efficient ML models ⚙️
Building stronger ML intuition, one concept at a time 🚀
@krishnaik06
#MachineLearning #PCA #DataScience #Python #AI

[Yugansh Agarwal @yuganshaga_dev]

Revised Exploratory Data Analysis (EDA) 📊
Practiced on Flight Price and Google Play Store datasets.
Explored patterns, cleaned data, visualized insights, and strengthened feature engineering skills ⚙️
@krishnaik06
#DS #EDA #Python #ML

[Yugansh Agarwal @yuganshaga_dev]

Revised Feature Engineering today ⚙️📊
Covered handling missing values, dealing with imbalanced datasets, and encoding categorical features for ML models🤖
Strengthening the data preprocessing skills that power better predictions🚀
@krishnaik06
#ML #DS #FeatureEngineering #Python