Publish artifacts for support-routing-ml-20260801

README.md

@@ -0,0 +1,59 @@
+---
+license: cc-by-4.0
+language:
+- en
+pretty_name: Applied ML Support Router Synthetic Evaluation Set
+size_categories:
+- n<1K
+task_categories:
+- text-classification
+tags:
+- synthetic
+- applied-machine-learning
+- evaluation
+- text-classification
+- zero-shot-classification
+- sentence-similarity
+- summarization
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train.jsonl
+  - split: test
+    path: data/test.jsonl
+---
+
+# Applied ML Support Router Synthetic Dataset
+
+## Summary
+
+This dataset contains 14 training examples and 4
+held-out examples for **Support operations need reproducible routing models that expose confidence and defer uncertain cases.**
+
+Every record is synthetic and includes:
+
+- `input`: query, event, or feature description
+- `label`: expected class, route, relation, or evidence category
+- `context`: synthetic supporting context
+- `source`: fictional source identifier
+- `variant`: generation pattern
+- `synthetic`: always `true`
+
+## Uses
+
+- Reproducible unit and integration tests
+- Baseline model training
+- Evaluation harness development
+- Schema and architecture demonstrations
+
+## Limitations
+
+Synthetic tickets do not represent every user population or language. Production training data needs consent and bias analysis.
+
+This dataset does not represent real users, patients, customers, production
+traffic, or licensed media. It must not be presented as real-world evidence.
+
+## Related Model
+
+[RKB109/support-routing-ml-20260801-model](https://huggingface.co/RKB109/support-routing-ml-20260801-model)

data/test.jsonl

@@ -0,0 +1,4 @@
+{"id":"support-routing-ml-2-1","input":"The API returns a timeout after deployment","label":"technical","context":"Technical failure affecting an integration.","source":"ticket-02","variant":"direct","synthetic":true}
+{"id":"support-routing-ml-3-2","input":"In an operations review, I cannot reset my password or access my account","label":"account-access","context":"Authentication and account recovery request.","source":"ticket-03","variant":"operations","synthetic":true}
+{"id":"support-routing-ml-4-3","input":"For an evaluation case, I received a suspicious login notification","label":"security","context":"Potential account compromise or unauthorized access.","source":"ticket-04","variant":"evaluation","synthetic":true}
+{"id":"support-routing-ml-6-1","input":"The SDK fails to parse the response body","label":"technical","context":"Client library behavior requires engineering review.","source":"ticket-06","variant":"direct","synthetic":true}

data/train.jsonl

@@ -0,0 +1,14 @@
+{"id":"support-routing-ml-1-1","input":"I was charged twice for the same subscription","label":"billing","context":"Duplicate charge investigation request.","source":"ticket-01","variant":"direct","synthetic":true}
+{"id":"support-routing-ml-1-2","input":"In an operations review, I was charged twice for the same subscription","label":"billing","context":"Duplicate charge investigation request.","source":"ticket-01","variant":"operations","synthetic":true}
+{"id":"support-routing-ml-1-3","input":"For an evaluation case, I was charged twice for the same subscription","label":"billing","context":"Duplicate charge investigation request.","source":"ticket-01","variant":"evaluation","synthetic":true}
+{"id":"support-routing-ml-2-2","input":"In an operations review, The API returns a timeout after deployment","label":"technical","context":"Technical failure affecting an integration.","source":"ticket-02","variant":"operations","synthetic":true}
+{"id":"support-routing-ml-2-3","input":"For an evaluation case, The API returns a timeout after deployment","label":"technical","context":"Technical failure affecting an integration.","source":"ticket-02","variant":"evaluation","synthetic":true}
+{"id":"support-routing-ml-3-1","input":"I cannot reset my password or access my account","label":"account-access","context":"Authentication and account recovery request.","source":"ticket-03","variant":"direct","synthetic":true}
+{"id":"support-routing-ml-3-3","input":"For an evaluation case, I cannot reset my password or access my account","label":"account-access","context":"Authentication and account recovery request.","source":"ticket-03","variant":"evaluation","synthetic":true}
+{"id":"support-routing-ml-4-1","input":"I received a suspicious login notification","label":"security","context":"Potential account compromise or unauthorized access.","source":"ticket-04","variant":"direct","synthetic":true}
+{"id":"support-routing-ml-4-2","input":"In an operations review, I received a suspicious login notification","label":"security","context":"Potential account compromise or unauthorized access.","source":"ticket-04","variant":"operations","synthetic":true}
+{"id":"support-routing-ml-5-1","input":"My invoice has the wrong tax amount","label":"billing","context":"Invoice calculation or billing documentation issue.","source":"ticket-05","variant":"direct","synthetic":true}
+{"id":"support-routing-ml-5-2","input":"In an operations review, My invoice has the wrong tax amount","label":"billing","context":"Invoice calculation or billing documentation issue.","source":"ticket-05","variant":"operations","synthetic":true}
+{"id":"support-routing-ml-5-3","input":"For an evaluation case, My invoice has the wrong tax amount","label":"billing","context":"Invoice calculation or billing documentation issue.","source":"ticket-05","variant":"evaluation","synthetic":true}
+{"id":"support-routing-ml-6-2","input":"In an operations review, The SDK fails to parse the response body","label":"technical","context":"Client library behavior requires engineering review.","source":"ticket-06","variant":"operations","synthetic":true}
+{"id":"support-routing-ml-6-3","input":"For an evaluation case, The SDK fails to parse the response body","label":"technical","context":"Client library behavior requires engineering review.","source":"ticket-06","variant":"evaluation","synthetic":true}

project.json

@@ -0,0 +1,52 @@
+{
+  "name": "Applied ML Support Router",
+  "problem": "Support operations need reproducible routing models that expose confidence and defer uncertain cases.",
+  "domain": "applied-machine-learning",
+  "architecture": "classifier",
+  "hugging_face_tasks": [
+    "text-classification",
+    "zero-shot-classification",
+    "sentence-similarity",
+    "summarization"
+  ],
+  "recommended_stack": [
+    "FastAPI for prediction and feedback endpoints",
+    "scikit-learn or LightGBM for production baselines",
+    "Sentence Transformers for semantic fallback",
+    "MLflow for experiments and model registry",
+    "PostgreSQL for labels and human feedback",
+    "Evidently-style drift and quality monitoring"
+  ],
+  "real_world_data_sources": [
+    {
+      "name": "GitHub Issues API",
+      "url": "https://api.github.com/repos/pytorch/pytorch/issues?state=open&per_page=10",
+      "purpose": "Real technical support-style issue text"
+    },
+    {
+      "name": "Stack Exchange API",
+      "url": "https://api.stackexchange.com/2.3/questions?pagesize=10&order=desc&sort=activity&site=stackoverflow",
+      "purpose": "Real developer questions for weak-label experiments"
+    }
+  ],
+  "job_description_skills": [
+    "Applied NLP classification and confidence calibration",
+    "Human feedback loops and active learning",
+    "Model registry, drift monitoring, and retraining",
+    "Feature, label, and evaluation pipeline design",
+    "Business-aware automation and escalation metrics"
+  ],
+  "impact_targets": [
+    "Reach macro F1 >= 0.85 on a reviewed support set",
+    "Automate >= 60% of tickets at >= 0.90 precision",
+    "Route low-confidence cases to human review",
+    "Detect label and feature drift before SLA impact"
+  ],
+  "baseline_evaluation": {
+    "test_examples": 4,
+    "accuracy": 1,
+    "synthetic_evaluation": true
+  },
+  "estimated_delivery": "8-12 weeks for one engineer",
+  "generated_baseline_is_production_ready": false
+}

sources.json

@@ -0,0 +1,12 @@
+[
+  {
+    "name": "GitHub Issues API",
+    "url": "https://api.github.com/repos/pytorch/pytorch/issues?state=open&per_page=10",
+    "purpose": "Real technical support-style issue text"
+  },
+  {
+    "name": "Stack Exchange API",
+    "url": "https://api.stackexchange.com/2.3/questions?pagesize=10&order=desc&sort=activity&site=stackoverflow",
+    "purpose": "Real developer questions for weak-label experiments"
+  }
+]