Fix tier matching: use client tier to pick correct complexity variant
- Doc parser now extracts tier labels (Tier A, A, Gold, etc.) per asset - Matching uses tier to find the correct GMAL complexity variant: - Claude matches to the GMAL family (asset type) - Post-match lookup: (asset_name + target_complexity_level) finds exact variant - e.g. "Banner - Tier A" with A=Complex → finds Complex variant by asset_name query - Tier hint passed to Claude prompt for better matching - No blind expansion - only the tier-appropriate GMAL is matched - Expand to Tiers button still available for when client doesn't specify tiers Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
DJP
parent
668ea44ea2
commit
3cb1973f57
3 changed files with 60 additions and 7 deletions
|
|
@ -63,6 +63,7 @@ async def _background_parse(project_id: int, filename: str, text: str, metadata:
|
|||
project_id=project_id,
|
||||
raw_name=item.get("name", "Unknown"),
|
||||
raw_description=item.get("description", ""),
|
||||
client_tier=item.get("tier", "") or None,
|
||||
volume=volume,
|
||||
sort_order=idx + 1,
|
||||
)
|
||||
|
|
|
|||
|
|
@ -108,14 +108,18 @@ Guidelines:
|
|||
- Be generous with scoring when the match is semantically correct even if the naming differs."""
|
||||
|
||||
|
||||
def _match_single_asset(client_asset_name, client_asset_desc, volume, catalog_text, num_assets):
|
||||
def _match_single_asset(client_asset_name, client_asset_desc, volume, catalog_text, num_assets, tier_hint=""):
|
||||
"""Run a single match call to Claude (synchronous, for use in thread pool)."""
|
||||
tier_instruction = ""
|
||||
if tier_hint:
|
||||
tier_instruction = f"\nCLIENT TIER: {tier_hint} — match to the {tier_hint} complexity variant if one exists.\n"
|
||||
|
||||
user_msg = f"""Match this client asset to the best GMAL equivalent(s):
|
||||
|
||||
CLIENT ASSET:
|
||||
Name: {client_asset_name}
|
||||
Description: {client_asset_desc or 'No description provided'}
|
||||
Volume: {volume}
|
||||
Volume: {volume}{tier_instruction}
|
||||
|
||||
FULL GMAL CATALOG ({num_assets} assets):
|
||||
{catalog_text}"""
|
||||
|
|
@ -143,9 +147,27 @@ async def match_client_assets(
|
|||
"""
|
||||
_clear_cancel(project_id)
|
||||
|
||||
# Load project tier mapping if set
|
||||
import json as _json
|
||||
from app.models.project import Project
|
||||
proj_result = await db.execute(select(Project).where(Project.id == project_id))
|
||||
project = proj_result.scalar_one_or_none()
|
||||
tier_config = {}
|
||||
if project and project.tier_mapping:
|
||||
try:
|
||||
tier_config = _json.loads(project.tier_mapping)
|
||||
except _json.JSONDecodeError:
|
||||
pass
|
||||
# Build tier label -> complexity_level map
|
||||
tier_to_complexity = {}
|
||||
complexity_map = {"simple": 1, "medium": 2, "mid": 2, "complex": 3}
|
||||
for t in tier_config.get("tiers", []):
|
||||
tier_to_complexity[t["label"].lower()] = complexity_map.get(t["complexity"].lower(), 2)
|
||||
|
||||
# Snapshot client asset data before any commits (ORM objects expire after commit)
|
||||
asset_snapshots = [
|
||||
{"id": ca.id, "raw_name": ca.raw_name, "raw_description": ca.raw_description, "volume": ca.volume}
|
||||
{"id": ca.id, "raw_name": ca.raw_name, "raw_description": ca.raw_description,
|
||||
"volume": ca.volume, "client_tier": getattr(ca, 'client_tier', None)}
|
||||
for ca in client_assets
|
||||
]
|
||||
|
||||
|
|
@ -155,6 +177,8 @@ async def match_client_assets(
|
|||
)
|
||||
all_gmals = result.scalars().all()
|
||||
gmal_by_id = {g.gmal_id: g for g in all_gmals}
|
||||
# Lookup by (asset_name, complexity_level) for tier variant resolution
|
||||
gmal_by_name_complexity = {(g.asset_name, g.complexity_level): g for g in all_gmals if g.asset_name}
|
||||
|
||||
# Build compact catalog once - reused for every match call
|
||||
catalog_text = _format_compact_catalog(all_gmals)
|
||||
|
|
@ -223,6 +247,15 @@ async def match_client_assets(
|
|||
for snap in batch:
|
||||
if _is_cancelled(project_id):
|
||||
break
|
||||
# Build tier hint for matching
|
||||
tier_hint = ""
|
||||
client_tier = snap.get("client_tier") or ""
|
||||
if client_tier and tier_to_complexity:
|
||||
complexity_level = tier_to_complexity.get(client_tier.lower())
|
||||
if complexity_level:
|
||||
complexity_names = {1: "Simple", 2: "Medium", 3: "Complex"}
|
||||
tier_hint = complexity_names.get(complexity_level, "")
|
||||
|
||||
future = loop.run_in_executor(
|
||||
executor,
|
||||
_match_single_asset,
|
||||
|
|
@ -231,6 +264,7 @@ async def match_client_assets(
|
|||
snap["volume"],
|
||||
catalog_text,
|
||||
len(all_gmals),
|
||||
tier_hint,
|
||||
)
|
||||
futures.append((snap, future))
|
||||
|
||||
|
|
@ -263,12 +297,24 @@ async def match_client_assets(
|
|||
logger.warning(f"Claude returned unknown GMAL ID: {m['gmal_id']}")
|
||||
continue
|
||||
|
||||
# If client asset has a tier, find the correct complexity variant
|
||||
final_gmal = gmal
|
||||
client_tier = snap.get("client_tier") or ""
|
||||
if client_tier and tier_to_complexity and gmal.asset_name:
|
||||
target_complexity = tier_to_complexity.get(client_tier.lower())
|
||||
if target_complexity and gmal.complexity_level != target_complexity:
|
||||
# Find the sibling at the right complexity
|
||||
variant = gmal_by_name_complexity.get((gmal.asset_name, target_complexity))
|
||||
if variant:
|
||||
logger.info(f"Tier remap: '{snap['raw_name']}' ({client_tier}) → {variant.gmal_id} ({variant.complexity_name}) instead of {gmal.gmal_id}")
|
||||
final_gmal = variant
|
||||
|
||||
match = Match(
|
||||
client_asset_id=snap["id"],
|
||||
gmal_asset_id=gmal.id,
|
||||
gmal_asset_id=final_gmal.id,
|
||||
confidence=MatchConfidence(m["confidence"]),
|
||||
confidence_score=m.get("confidence_score"),
|
||||
ai_reasoning=m.get("reasoning"),
|
||||
ai_reasoning=m.get("reasoning") + (f" [Tier {client_tier} → {final_gmal.complexity_name}]" if final_gmal != gmal else ""),
|
||||
caveat_text=m.get("caveats"),
|
||||
is_selected=(rank == 1 and auto_select),
|
||||
rank=rank,
|
||||
|
|
|
|||
|
|
@ -40,8 +40,12 @@ EXTRACT_TOOLS = [
|
|||
"type": "integer",
|
||||
"description": "Number of this asset needed (default 1 if not specified)"
|
||||
},
|
||||
"tier": {
|
||||
"type": "string",
|
||||
"description": "The client's tier/complexity label if specified (e.g. 'Tier A', 'A', 'Gold', '1', 'Premium'). Leave empty string if no tier is specified."
|
||||
},
|
||||
},
|
||||
"required": ["name", "description", "complexity_hint", "volume"],
|
||||
"required": ["name", "description", "complexity_hint", "volume", "tier"],
|
||||
},
|
||||
},
|
||||
},
|
||||
|
|
@ -58,8 +62,10 @@ For each asset, provide:
|
|||
- description: What this asset involves based on the document context. Include format, size, channel, and any other relevant details.
|
||||
- complexity_hint: Your best estimate of complexity (simple/medium/complex) based on the description. Use "unknown" if unclear.
|
||||
- volume: How many of this asset are needed. Default to 1 if not specified.
|
||||
- tier: If the client specifies a tier, grade, or complexity label for this asset (e.g. "Tier A", "A", "Gold", "Premium", "1"), include it exactly as written. If the document has columns like A/B/C or Tier 1/2/3, extract those labels. Leave empty string if no tier is specified.
|
||||
|
||||
Be thorough - extract every distinct asset type mentioned. If the same asset appears at different complexity levels, list them separately.
|
||||
Be thorough - extract every distinct asset type mentioned. If the same asset appears at different tiers or complexity levels, list them as SEPARATE entries with their respective tier labels.
|
||||
Do NOT combine different asset types into one entry.
|
||||
Do NOT combine different asset types into one entry."""
|
||||
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue