Muhammad Haris

A meaningful discussion on how technology is transforming the way we learn, work, and innovate.

AI in Pakistan Jobs, Education & Reality | Part - 2 In this exclusive podcast episode, we sit down with a Sr. Machine Learning Engineer to explore the realities of Artificial Intelligence from its foundations ...

Had an insightful conversation about the impact of AI on the new generation and the opportunities it brings.

AI in Pakistan Jobs, Education & Reality | Part -1 In this exclusive podcast episode, we sit down with a Sr. Machine Learning Engineer to explore the realities of Artificial Intelligence from its foundations ...

Alhamdulillah, I’ve completed my MPhil in Data Science from Information Technology University . Even before graduation, I had the opportunity to apply AI and Data Science to real industry challenges.
One lesson I carry forward: Strategy without ex*****on is just an idea, and ex*****on without vision is just work.
I’m deeply grateful to my parents for their endless support. The journey continues.

الحمدللہ، میں نے انفارمیشن ٹیکنالوجی یونیورسٹی سے ایم فل ان ڈیٹا سائنس مکمل کر لیا ہے۔ گریجویشن سے پہلے ہی مجھے اے آئی اور ڈیٹا سائنس کو حقیقی صنعتی مسائل پر لاگو کرنے کا موقع ملا۔
ایک سبق جو میں ہمیشہ ساتھ لے کر چلوں گا: حکمتِ عملی بغیر عمل صرف ایک خیال ہے، اور عمل بغیر وژن صرف کام۔
میں اپنے والدین کا دل سے شکر گزار ہوں۔

What I’ve Learned from Working with RAG Systems

Over the past few years of developing and experimenting with Retrieval-Augmented Generation (RAG) systems, I’ve realized that while the vanilla RAG approach looks effective in theory, it faces significant limitations in real-world enterprise environments.

1. The Retrieval Problem
In practice, the quality of retrieval defines the performance of the entire system. Even with a strong Large Language Model (LLM), if the retriever fails to fetch the correct context, the generated output will be incomplete or inaccurate. Common issues include partial or irrelevant context retrieval, poor chunking strategies when information is distributed across sections, and overreliance on vector similarity that ignores logical and contextual relationships. The retrieval layer often becomes the bottleneck of the entire RAG pipeline.

2. Complex Data Structures
Another challenge appears when dealing with structured or semi-structured data. Vanilla RAG performs well on plain text but struggles with complex sources such as financial reports, legal documents, and enterprise knowledge bases. These data types include interconnected entities, hierarchies, and references that cannot be represented effectively in a flat vector store. This fragmentation causes loss of relationships that define the actual meaning of the content. Understanding data relationships is as critical as understanding the text itself.

Key Takeaway
To overcome these limitations, we need to move beyond simple text-based retrieval and adopt Structured RAG approaches. By integrating Knowledge Graphs and multi-component retrieval pipelines, we can capture both semantic and relational dimensions of information, resulting in more accurate and context-aware outputs.

In upcoming posts, I’ll share how Knowledge Graph–augmented RAG systems can address these issues and enable more intelligent retrieval pipelines.

میں Irfan Malik کی ٹیم کا ایک حصہ ہونے پر فخر محسوس کرتا ہوں، اور میں نے ان سے بہت کچھ سیکھا ہے۔ انہوں نے مجھے سکھایا ہے کہ کبھی بھی ہمت نہیں ہارنی چاہیے، اور اپنے مقاصد کے حصول کے لیے ہمیشہ کوشاں رہنا چاہیے۔

میں ان کی اس پوسٹ کی تائید کرتا ہوں، اور میں یقین رکھتا ہوں کہ اگر ہم سب مل کر کام کریں تو ہم اپنے تمام خوابوں کو پورا کر سکتے ہیں۔

"Our greatest weakness lies in giving up. The most certain way to succeed is always to try just one more time." - Thomas Edison