From ae40d9d8a36f7c48df2ececff7b5f3436589ca75 Mon Sep 17 00:00:00 2001 From: Harshit Chouhan Date: Thu, 12 Nov 2026 12:13:49 +0530 Subject: [PATCH 1/2] Add article: Semantic Execution Layer for Enterprise AI --- weekly/semantic-execution-layer-architecture.yml | 12 ++++++++++++ 1 file changed, 12 insertions(+) create mode 100644 weekly/semantic-execution-layer-architecture.yml diff --git a/weekly/semantic-execution-layer-architecture.yml b/weekly/semantic-execution-layer-architecture.yml new file mode 100644 index 0000000..ee27633 --- /dev/null +++ b/weekly/semantic-execution-layer-architecture.yml @@ -0,0 +1,12 @@ +article: + author: + linkedin: "https://www.linkedin.com/in/h-chouhan/" + name: "Harshit Chouhan" + twitter: "https://x.com/chouhanharshit1" + link: "https://colrows.com/whitepapers/semantic-execution-layer-for-enterprise-ai/" + review: "A technical look at why text-to-SQL agents fail on real enterprise schemas, and a compiler-style answer to it. The paper treats business meaning as a typed, versioned knowledge graph and moves the query through seven staged phases before any SQL runs: parsing, semantic binding, join-path proof, constraint solving, logical planning, cost-based physical planning, and dialect generation. Useful details for data engineers: bounded BFS with grain and fan-out pruning to prove a join path instead of guessing it, drift detection with KL-divergence, KS statistics and tree edit distance on schema snapshots, and access policy applied as a scope subgraph at compile time rather than masking after execution." + reviewer: + linkedin: "" + name: "" + twitter: "" + tags: "semantic layer, knowledge graph, query compilation, join path, schema drift, data governance, text-to-SQL, AI agents" From c59270c97d9e40407a7d57c14ce6e95ff4568085 Mon Sep 17 00:00:00 2001 From: Harshit Chouhan Date: Thu, 12 Nov 2026 12:15:28 +0530 Subject: [PATCH 2/2] Update semantic-execution-layer-architecture.yml --- weekly/semantic-execution-layer-architecture.yml | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/weekly/semantic-execution-layer-architecture.yml b/weekly/semantic-execution-layer-architecture.yml index ee27633..f395cff 100644 --- a/weekly/semantic-execution-layer-architecture.yml +++ b/weekly/semantic-execution-layer-architecture.yml @@ -6,7 +6,7 @@ article: link: "https://colrows.com/whitepapers/semantic-execution-layer-for-enterprise-ai/" review: "A technical look at why text-to-SQL agents fail on real enterprise schemas, and a compiler-style answer to it. The paper treats business meaning as a typed, versioned knowledge graph and moves the query through seven staged phases before any SQL runs: parsing, semantic binding, join-path proof, constraint solving, logical planning, cost-based physical planning, and dialect generation. Useful details for data engineers: bounded BFS with grain and fan-out pruning to prove a join path instead of guessing it, drift detection with KL-divergence, KS statistics and tree edit distance on schema snapshots, and access policy applied as a scope subgraph at compile time rather than masking after execution." reviewer: - linkedin: "" - name: "" - twitter: "" + linkedin: "https://www.linkedin.com/in/h-chouhan/" + name: "Harshit Chouhan" + twitter: "https://x.com/chouhanharshit1" tags: "semantic layer, knowledge graph, query compilation, join path, schema drift, data governance, text-to-SQL, AI agents"