GraphRAG - A new and Better Approach for Retrieval

[manas95826]

Talk title

GraphRAG - A new and Better Approach for Retrieval

Short talk description

GraphRAG is an advanced retrieval architecture that combines semantic vector search with knowledge graph traversal to improve contextual reasoning, multi-hop retrieval, and connected knowledge understanding. Using Qdrant for semantic retrieval and graph databases like Neo4j for relationship traversal, GraphRAG enables more accurate and context-aware AI systems compared to traditional RAG pipelines.

Long talk description

GraphRAG is a next-generation retrieval architecture that extends traditional Retrieval-Augmented Generation (RAG) by combining semantic vector search with knowledge graph-based reasoning. While traditional RAG systems retrieve semantically similar chunks using embeddings, GraphRAG enhances this process by introducing entities, relationships, and graph traversal to retrieve connected and contextually related information.

In a typical GraphRAG pipeline, documents are first chunked and processed through two parallel workflows: embeddings are generated and stored in vector databases like Qdrant for semantic retrieval, while entities and relationships are extracted using LLMs or graph transformers and stored in graph databases like Neo4j. During retrieval, Qdrant handles fast ANN-based semantic search, while the graph layer performs multi-hop traversal across connected entities. The retrieved semantic and graph contexts are then fused and passed to the LLM for grounded reasoning and answer generation.

This architecture is especially useful in domains where relationships matter deeply, such as cybersecurity, enterprise knowledge systems, supply chain analysis, biomedical research, fraud detection, and AI agents. GraphRAG improves contextual understanding, connected reasoning, and retrieval quality by moving beyond isolated chunk retrieval into structured knowledge-aware retrieval systems.

What format do you have in mind?

Talk (20-25 minutes + Q&A)

Talk outline / Agenda

• Introduction to GraphRAG
• Limitations of Traditional RAG
• What Knowledge Graphs Add to Retrieval
• Core Components of a GraphRAG Pipeline
• Role of Qdrant in GraphRAG
• Semantic Search vs Graph Traversal
• Where GraphRAG Shines
• Challenges & The Modern Hybrid Architecture
• Key Takeaways & Discussion

Key takeaways

• Traditional RAG retrieves similar chunks; GraphRAG retrieves connected knowledge.
• Qdrant enables scalable semantic retrieval through fast ANN vector search.
• Knowledge graphs improve multi-hop reasoning and contextual understanding.
• Modern AI systems increasingly combine vectors, graphs, and memory architectures.
• Hybrid retrieval pipelines are becoming the foundation of next-generation AI agents and enterprise RAG systems.

What domain would you say your talk falls under?

Core Python

Duration (including Q&A)

40min

Prerequisites and preparation

No response

Resources and references

No response

Link to slides/demos (if available)

https://drive.google.com/file/u/1/d/1tx311jmDhljaFdUh7n7AQEZ74xacIOKJW166gG7qYpQ/view?pli=1

Twitter/X handle (optional)

No response

LinkedIn profile (optional)

https://linkedin.com/in/themanasai

Profile picture URL (optional)

No response

Speaker bio

Manas Chopra is a Developer Relations & Community Manager at Qdrant, focused on vector search, RAG systems, AI infrastructure, and developer ecosystem growth. He is also the co-founder of Geek Room, one of India’s largest developer communities, where he has led large-scale hackathons, workshops, and AI-focused initiatives reaching over 100K+ developers.

Reach out to me through my linkedin or manas.chopra@qdrant.com

Availability

23/05/2026

Accessibility & special requirements

No response

Speaker checklist

• I have read and understood the PyDelhi guidelines for submitting proposals and giving talks
• I have read and acknowledged the PyDelhi accessibility guidelines and will ensure my presentation materials (slides, videos, demos) follow these recommendations
• I will make my talk accessible to all attendees and will proactively ask for any accommodations or special requirements I might need
• I agree to share slides, code snippets, and other materials used during the talk with the community
• I will follow PyDelhi's Code of Conduct and maintain a welcoming, inclusive environment throughout my participation
• I understand that PyDelhi meetups are community-centric events focused on learning, knowledge sharing, and networking, and I will respect this ethos by not using this platform for self-promotion or hiring pitches during my presentation, unless explicitly invited to do so by means of a sponsorship or similar arrangement
• If the talk is recorded by the PyDelhi team, I grant permission to release the video on PyDelhi's YouTube channel under the CC-BY-4.0 license, or a different license of my choosing if I am specifying it in my proposal or with the materials I share

Additional comments

No response

[github-actions]

Thank you for submitting your proposal! 🎉

Hi @manas95826! Thank you for taking the time to share your learnings with the PyDelhi community! 👋

The PyDelhi volunteer team (@pydelhi/reviewers) has been notified of your submission and will review it as soon as possible. We appreciate your patience during this process.

What happens next?

1. Review process: The team will review your proposal as soon as possible.
2. Feedback: Reviewers may reach out to seek clarifications or suggestions to help strengthen your proposal, to ensure it aligns with our guidelines, and to make it more engaging for the audience. Please keep an eye on your notifications!
3. Confirmation: Once your proposal is formally accepted, the event logistics will be coordinated with you, including scheduling and any necessary preparations.
4. Support: First-time speakers can seek mentorship and practice sessions if needed, subject to the availability of the team. Please let us know!

Important reminders

• Please make sure you have read and understood the guidelines for submitting proposals and delivering talks, as well as our accessibility guidelines.
• All talks/sessions are licensed under a Creative Commons Attribution-Non Commercial-ShareAlike 4.0 International License, unless you state otherwise in your proposal above. This means that your talk can be shared and adapted for non-commercial purposes, as long as appropriate credit is given and any derivative works are shared under the same license.
• Please follow PyDelhi's Code of Conduct and maintain a welcoming, inclusive environment throughout your participation.

Questions?

Feel free to:

• Join the PyDelhi WhatsApp community
• Join the PyDelhi Telegram group
• Send us an email at contact@pydelhi.org
• Write to us in a comment below this message! :)

Thank you again for contributing to the Python community in Delhi! 🐍❤️

---

Improve this questionnaire: Did you find a bug, or do you have suggestions? Please feel free to add them as a comment below or open a separate issue as you prefer. We welcome contributions to improve this proposal form and the overall process!

This is an automated message. The review team will be with you shortly!

This proposal form was inspired by and enhanced with ideas from BangPypers. Thank you to the BangPypers community for sharing best practices!

[Arko-Sengupta]

Hi @manas95826,

Thank you for submitting your proposal.

We found the topic interesting and believe it covers valuable concepts. However, we would recommend a few enhancements before moving forward. The current content appears to be geared more toward a beginner-level audience, whereas the majority of attendees for this session are expected to be experienced professionals. As a result, the talk may feel somewhat introductory for the target audience.

Additionally, the slides seem to rely heavily on AI-generated content. It would be great if the material could be made more practical and grounded in real-world examples, use cases, lessons learned, or implementation experiences.

For a more professional audience, you may consider expanding on advanced topics such as:

• Graph traversal techniques and optimization strategies
• HNSW (Hierarchical Navigable Small World) indexing and search
• Fast retrieval mechanisms and performance tuning
• Techniques for improving retrieval accuracy and relevance
• Scalability considerations and production best practices

We believe incorporating these aspects would make the session more impactful and better aligned with the expectations of the audience.

Looking forward to seeing the revised version.

Best regards,
Arko Sengupta