Checkpoint frontend UI/UX overhaul and ingestion embed benchmark work

apps/ai-service/adapters/embedding.py

@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-import hashlib
-import math
 from typing import Any
 
 from rag.ports import QueryEmbeddingUnavailable
@@ -120,59 +118,3 @@ class BedrockCohereQueryEmbedder:
                 f"expected {self._dimensions}"
             )
         return values
-
-
-class SectionOnlyQueryEmbedder:
-    """Refuses to embed, confining retrieval to the route that is verified.
-
-    The section route resolves the question's attribute to a `section_key` and
-    filters on it; no vector is involved, and it measured 16/16 on the
-    human-written golden questions on 2026-08-04. Similarity measured 0.544 and
-    its provider is currently revoked.
-
-    Declining locally and immediately is better than the two alternatives it
-    replaces: a `cohere-v4` round-trip spends the boto3 retry budget before
-    failing, and `LocalHashQueryEmbedder` searches a SHA-256 vector against a
-    Cohere collection, which returns confident and meaningless hits.
-    """
-
-    def __init__(self, dimensions: int) -> None:
-        if dimensions <= 0:
-            raise ValueError("dimensions must be positive")
-        self._dimensions = dimensions
-
-    @property
-    def dimensions(self) -> int:
-        return self._dimensions
-
-    def embed_query(self, text: str) -> list[float]:  # noqa: ARG002
-        # The text is irrelevant: this embedder exists to refuse, not to embed.
-        raise QueryEmbeddingUnavailable(
-            "no query embedding provider is enabled; set EMBEDDING_PROVIDER to "
-            "use the similarity fallback"
-        )
-
-
-class LocalHashQueryEmbedder:
-    """Deterministic local plumbing probe; not a semantic retrieval model."""
-
-    def __init__(self, dimensions: int) -> None:
-        if dimensions <= 0:
-            raise ValueError("dimensions must be positive")
-        self._dimensions = dimensions
-
-    @property
-    def dimensions(self) -> int:
-        return self._dimensions
-
-    def embed_query(self, text: str) -> list[float]:
-        vector = [0.0] * self._dimensions
-        for token in text.casefold().split():
-            digest = hashlib.sha256(token.encode("utf-8")).digest()
-            index = int.from_bytes(digest[:4], "big") % self._dimensions
-            sign = 1.0 if digest[4] & 1 else -1.0
-            vector[index] += sign
-        norm = math.sqrt(sum(value * value for value in vector))
-        if norm == 0:
-            return vector
-        return [value / norm for value in vector]