Swapnil Bandwal Foundry IQ Badge Request

[swapllbandwal]

Curriculum Completion Validation

• I verify that I have completed all 3 Foundry IQ episode cookbooks end-to-end.

Ep 1 — Unlocking Knowledge for Agents: solution verification

• I verify that I have uploaded a screenshot of my Episode 1 final notebook output with my GitHub username or Azure resource name shown into the field below.

Ep 1 — fork URL

https://github.com/swapllbandwal/iq-series/tree/main/1-Foundry-IQ-Unlocking-Knowledge-for-Agents

Please note : All steps in the episode were completed successfully, and the expected outputs were achieved.
However, while attempting to push the notebook cell outputs to the remote repository, an error occurred a repository rule violation.
Please find the attached screenshot for reference.

Ep 1 — final output screenshot

Foundry iq - Episode_1.docx

Ep 2 — Building the Data Pipeline with Knowledge Sources: solution verification

• I verify that I have uploaded a screenshot of my Episode 2 final notebook output with my GitHub username or Azure resource name shown into the field below.

Ep 2 — fork URL

https://github.com/microsoft/iq-series/tree/main/2-Foundry-IQ-Building-the-Data-Pipeline-with-Knowledge-Sources/cookbook

Please note : All steps in the episode were completed successfully, and the expected outputs were achieved.
However, while attempting to push the notebook cell outputs to the remote repository, an error occurred a repository rule violation.
Please find the attached screenshot for reference.

Ep 2 — final output screenshot

Foundry iq - Episode_2.docx

Ep 3 — Querying the Multi-Source AI Knowledge Bases: solution verification

• I verify that I have uploaded a screenshot of my Episode 3 final notebook output with my GitHub username or Azure resource name shown into the field below.

Ep 3 — fork URL

https://github.com/microsoft/iq-series/tree/main/3-Foundry-IQ-Querying-the-Multi-Source-AI-Knowledge-Bases/cookbook

Please note : All steps in the episode were completed successfully, and the expected outputs were achieved.
However, while attempting to push the notebook cell outputs to the remote repository, an error occurred a repository rule violation.
Please find the attached screenshot for reference.

Ep 3 — final output screenshot

Foundry iq - Episode_3.docx

Episode Insights & Key Takeaways

Episode 1: Unlocking Knowledge for Agents
My Insight:
Agentic retrieval decomposes complex user queries into focused subqueries, runs them in parallel against a search index, reranks results semantically, and synthesizes a single grounded answer with inline citations , all in one API call. This eliminates the need for manual retrieval orchestration in agent code.

Episode 2: Building the Data Pipeline with Knowledge Sources
My Insight:
Knowledge Sources can wrap diverse data types—indexed search indexes, real-time Blob Storage ingestion pipelines, and live web content, within a single knowledge base. This means you don't need separate retrieval logic for each data source; the knowledge base orchestrates them transparently, enabling truly multi-source AI queries.

Episode 3: Querying the Multi-Source AI Knowledge Bases
My Insight:
Reasoning effort settings (minimal, low, medium) allow you to trade cost against answer quality in multi-source retrieval. Minimal effort skips LLM planning for fast queries; low effort adds planning; medium adds iterative refinement, giving teams fine-grained control over retrieval behavior without code changes.

Challenges or feedback

Feedback
The videos and cookbooks were incredibly insightful and engaging , best hands-on experience. The learning journey from RAG to Agentic RAG using Foundry IQ was seamless and well-structured. I'm excited to build real-world applications leveraging these capabilities.

Challenges : Overall I faced challenges at just 2 places :

1. Episode 3 - Minimal Reasoning Effort: The retrieval API doesn't support messages parameter during minimal retrieval; intent should be used instead. (Screenshot attached)

2. Git Push Restriction: Unable to push all episode notebook outputs to forked repo due to repository restrictions. (Screenshot attached)

Apart from these two issues, the overall experience was outstanding.
Thank you to the Global AI Community for this excellent learning initiative. Looking forward to earning the badge and tackling the next Azure AI challenges ahead! 👍

Badge form confirmation

• I verify that I have completed the Foundry IQ badge form.

[swapllbandwal]

Deployed Resources_Azure

Final Output_Episode 1 :

Final Output_Episode 2:

Final Output_Episode 3:
: