---
title: "Amoy Dutta (@amoydutta) | Hashnode"
description: "Amoy Dutta (@amoydutta) writes on Hashnode. Read their latest articles and follow their work."
url: "https://hashnode.com/@amoydutta"
---

@amoydutta

# Amoy Dutta

@amoydutta · Joined August 2025

### [Build Your Own Data Analysis AI Agent with smolagents and a Local LLM](https://blogs-amoydutta.hashnode.dev/build-your-own-data-analysis-ai-agent-with-smolagents-and-a-local-llm?utm_source=hashnode&utm_medium=feed)

Build a data analysis AI agent with smolagents, Python, and a local LLM. Explore CSV files, generate charts, and improve agent reliability.

14 min read · [blogs-amoydutta.hashnode.dev](https://blogs-amoydutta.hashnode.dev/build-your-own-data-analysis-ai-agent-with-smolagents-and-a-local-llm?utm_source=hashnode&utm_medium=feed)

[Oct 12](https://hashnode.com/posts/build-your-own-data-analysis-ai-agent-with-smolagents-and-a-local-llm/6acd1bbb6ee1ce179ae94fe1)

### [Can Jev Actually Sort Your IT Tickets? We Built a Baseline and Found Out](https://blogs-amoydutta.hashnode.dev/can-jev-actually-sort-your-it-tickets-we-built-a-baseline-and-found-out?utm_source=hashnode&utm_medium=feed)

A rigorous evaluation of Jev’s AI classifier, testing accuracy, calibration, labeling, and real-world performance against a TF-IDF + SVM baseline.

18 min read · [blogs-amoydutta.hashnode.dev](https://blogs-amoydutta.hashnode.dev/can-jev-actually-sort-your-it-tickets-we-built-a-baseline-and-found-out?utm_source=hashnode&utm_medium=feed)

[Oct 2](https://hashnode.com/posts/can-jev-actually-sort-your-it-tickets-we-built-a-baseline-and-found-out/6abfe3d7149b988bd69d9d18)

### [Explainable Recommendations in Practice: A Demo-Driven Approach](https://blogs-amoydutta.hashnode.dev/explainable-recommendations-in-practice-a-demo-driven-approach?utm_source=hashnode&utm_medium=feed)

A practical and mathematical walkthrough of building an explainable movie recommendation system using personas, collaborative filtering, and LLMs.

8 min read · [blogs-amoydutta.hashnode.dev](https://blogs-amoydutta.hashnode.dev/explainable-recommendations-in-practice-a-demo-driven-approach?utm_source=hashnode&utm_medium=feed)

[Feb 10](https://hashnode.com/posts/explainable-recommendations-in-practice-a-demo-driven-approach/698b6fb8ce66e4a119073284)

### [The End of the Black Box: Why the "Why" is the Future of Personalization](https://blogs-amoydutta.hashnode.dev/the-end-of-the-black-box-why-the-why-is-the-future-of-personalization?utm_source=hashnode&utm_medium=feed)

How LLM-generated explanations improve recommendation systems. Explore architecture, safety layers, performance trade-offs, and ROI for product teams.

8 min read · [blogs-amoydutta.hashnode.dev](https://blogs-amoydutta.hashnode.dev/the-end-of-the-black-box-why-the-why-is-the-future-of-personalization?utm_source=hashnode&utm_medium=feed)

[Jan 26](https://hashnode.com/posts/the-end-of-the-black-box-why-the-why-is-the-future-of-personalization/69775778df93d51483c90c35)

### [Building a Smarter Product Chatbot: What RAG Gets Right—and Where It Breaks](https://blogs-amoydutta.hashnode.dev/building-a-smarter-product-chatbot-what-rag-gets-rightand-where-it-breaks?utm_source=hashnode&utm_medium=feed)

A hands-on comparison of a prompt-engineered assistant and a RAG pipeline, with real failure modes, evaluation results, and practical takeaways.

11 min read · [blogs-amoydutta.hashnode.dev](https://blogs-amoydutta.hashnode.dev/building-a-smarter-product-chatbot-what-rag-gets-rightand-where-it-breaks?utm_source=hashnode&utm_medium=feed)

[Dec 22, 2025](https://hashnode.com/posts/building-a-smarter-product-chatbot-what-rag-gets-rightand-where-it-breaks/694969a019a9eb8a5131e133)