Files
Katon Minhas afb6d5185d Merged in feature/lesser-table-caching-refactor-hybrid (pull request #847)
Feature/lesser table caching refactor hybrid

* chore: Remove unused duplicate main.py from shared pipeline

* fix: Correct crosswalk paths in aarete_derived.py

* chore: Remove unused documentation files from fieldExtraction

* docs: Add documentation files to documentation folder

* docs: Update README with uv setup, expanded project structure, and branching conventions

* docs: Add uv installation steps with Ubuntu/WSL emphasis

* Enable prompt caching for all remaining LLM calls

- Add _INSTRUCTION() functions for: EXHIBIT_HEADER, EXHIBIT_LINKAGE,
  EXHIBIT_TITLE_MATCH, DATE_FIX, DERIVED_TERM_DATE, CHECK_PROVIDER_NAME_MATCH,
  SPECIAL_CASE_ASSIGNMENT
- Update all invoke_claude() calls in saas and clover pipelines to use
  cache=True with corresponding _INSTRUCTION() functions
- Add new instructions to get_cacheable_instructions() for cache warming
- Update tests for new instruction functions

Functions now using caching:
- prompt_exhibit_level
- prompt_exhibit_lesser (EXHIBIT_LEVEL_LESSER_OF)
- prompt_fee_schedule_breakout
- prompt_grouper_breakout
- prompt_special_case_assignment
- prompt_exhibit_linkage
- prompt_exhibit_header
- prompt_smart_chunked (ONE_TO_ONE templates)
- prompt_date_fix
- prompt_derived_term_date
- prompt_exhibit_title_match
- provider_name_match_check

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* Reorder

* feat: Add bcbs_promise client pipeline with OFFSET_TERM extraction

- Add new bcbs_promise client with HSC-based OFFSET_TERM field extraction
- Extract full paragraph text of offset/recoupment provisions from contracts
- Derive OFFSET_INDICATOR (Y/N) from OFFSET_TERM presence
- Fix reorder_columns to preserve extra columns not in COLUMN_ORDER
- Update QC/QA output path to outputs/qc_qa/

* fix: Update dev deps and test assertions for QC/QA output path

- Add pytest/pytest-mock to dev dependencies for mypy type checking
- Update test assertions to expect outputs/qc_qa instead of qa_qc_output

* style: Apply black formatting to prompt_templates.py

* Merge main, move scripts

* Archive some scripts

* update py version

* remove .py version file

* Remove ASCII characters

* Restore testbed code

* restore tracking

* Update testbed metrics

* Enable prompt caching for CODE_LAST_CHECK, FILL_BILL_TYPE, DUAL_LOB_CHECK, and GROUPER_BREAKOUT

- Add CODE_LAST_CHECK_INSTRUCTION() for service specificity classification
- Add FILL_BILL_TYPE_INSTRUCTION() for bill type code determination
- Add DUAL_LOB_CHECK_INSTRUCTION() for Medicare/Medicaid classification
- Update code_funcs.py to use caching for CODE_LAST_CHECK, FILL_BILL_TYPE, GROUPER_BREAKOUT
- Update postprocessing_funcs.py to use caching for DUAL_LOB_CHECK
- Add new instructions to get_cacheable_instructions() for cache warming
- Add unit tests for new instruction functions

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* Fix postprocessing_funcs to remove invalid columns

* Merge branch 'main' into feature/lesser-table-caching-refactor-hybrid

* Revert prompt caching changes from aed1b73c

* update formatting

* Update imports


Approved-by: Sha Brown
Approved-by: Praneel Panchigar
2026-01-26 16:52:55 +00:00

292 lines
9.9 KiB
Python

import json
import security
import streamlit as st
from streamlit_extras.add_vertical_space import add_vertical_space
import os
import streamlit as st
import pandas as pd
from io import StringIO
from datetime import datetime
import boto3
import util
import requests
from sf_conn import get_secret, save_to_sf
from io import StringIO, BytesIO
import time
from sf_conn import get_client_names, insert_upload_logs
from constants import USER_LIST
from util import logger
(REDIRECT_URI, create_batch_url, doczy_pipeline) = util.load_page_details(0)
user_list = USER_LIST
if "uploading" not in st.session_state:
st.session_state.uploading = False
if "upload_key" not in st.session_state:
st.session_state.upload_key = 0
if "file_list" not in st.session_state:
st.session_state.file_list = []
if "show_batchID" not in st.session_state:
st.session_state.show_batchID = False
if "landing_zone" not in st.session_state:
st.session_state.landing_zone = "contracts-landing-zone"
if "batch_id" not in st.session_state:
st.session_state.batch_id = "failed_cases"
if "client_bucket" not in st.session_state:
st.session_state.client_bucket = "default_bucket"
def insert_upload_logs(batch_id, client_name, file_name, upload_datetime, upload_user):
"""
Input: batch_id, client_name, file_name, upload_datetime, upload_user
Output: status of the insert query
"""
try:
return "Log inserted successfully"
except Exception as e:
return e
st.set_page_config(layout="wide")
# # # Sidebar contents
with st.sidebar:
st.title("Doczy.AI ™")
st.markdown(
"""
## About
This app extracts data from contracts
"""
)
add_vertical_space(15)
# st.write("Doczy")
# AARETE LOGO
x, y, z = st.columns([15, 2, 15])
with y:
st.image("aaretelogo.png")
hide_img_fs = """
<style>
button[title="View fullscreen"]{
visibility: hidden;}
</style>
"""
st.markdown(hide_img_fs, unsafe_allow_html=True)
_, c1 = st.columns([5, 1])
try:
util.setup_page(REDIRECT_URI)
except Exception as e:
st.write(f"SSO Failed = {e}")
st.session_state["user_info"] = {
"mail": "maamseek@aarete.com",
"displayName": "Mayank Aamseek",
} # RECOMMENDATION - Remove Line after development.
try:
c1.write(f"User: **{st.session_state.user_info['displayName']}**")
user_mail = st.session_state.user_info["mail"]
except (
KeyError
) as e: # RECOMMENDATION - Not Working as Expected. Unreachable Code because the above Exception handles everything
st.write("Session Expired.")
auth_url = security.get_auth_url(REDIRECT_URI)
st.markdown(
f"<a href='{auth_url}' target='_self'>Sign In</a>", unsafe_allow_html=True
)
st.stop()
s3_client = boto3.client(
"s3", # RECOMMENDATION - We can make use of AWS Access Key for enabling this in local testing. Will help a lot.
region_name="us-east-2",
)
client_list = [
"doczy-ai-client-1",
"Delaware First Health, Inc.",
"Community Health Choice, Inc",
"CareSource Network Partners LLC",
"HealthNet of Cali",
"Oklahoma Complete Health, Inc",
"HealthFirst",
"Molina Healthcare of TX",
"AvMed",
"Arizona Care1st",
"WellCare New Jersey",
]
# This is the list of client fetched from Snowflake
# TODO: Need to update the streamlit code to use the client names from this list
# And use the s3 paths to save the objects for the respective client
client_list, s3_paths = get_client_names()
logger.info(f"client_list={client_list}")
logger.info(f"s3_paths={s3_paths}")
client_s3_paths = dict(zip(client_list, s3_paths))
client_row = st.columns([0.1, 0.8])
with client_row[0]:
st.write("**Client Name**")
with client_row[1]:
client = st.selectbox(
"Client Name", (client_list), label_visibility="collapsed", index=None
)
client_bucket = client_s3_paths.get(client)
# to be deleted when client buckets are created
# client_bucket = 'doczyai-use2-d-cn1-s3-textract-processing-001'
file_row = st.columns([0.1, 0.8])
with file_row[0]:
st.write("**Upload Files**")
with file_row[1]:
file_list = st.file_uploader(
"Upload",
type=["docx", "tiff", "pdf"],
accept_multiple_files=True,
label_visibility="collapsed",
help="Only PDF, TIFF and DOCX file formats are supported.",
disabled=st.session_state.uploading,
key=st.session_state.upload_key,
)
add_vertical_space(2)
df = pd.DataFrame(columns=["Contract Name"])
df["Contract Name"] = file_list
file_names = []
buttons = st.columns([0.4, 0.4, 0.2])
def set_uploading_state():
if not client == None and not len(file_list) == 0:
st.session_state.file_list = file_list
st.session_state.upload_key += 1
st.session_state.uploading = True
with buttons[1]:
if st.button("Create Batch", on_click=set_uploading_state):
file_list = st.session_state.file_list
if client == None:
st.error("No Client Name Selected.")
elif len(file_list) == 0:
st.error("No Files Selected.")
else:
with st.spinner("Running..."):
myobj = {"client-bucket-name": client_bucket}
response = requests.post(create_batch_url, json=myobj)
if response.status_code >= 200 and response.status_code < 300:
try:
batch_id = json.loads(json.loads(response.text)["body"])[
"batch_id"
]
landing_zone = json.loads(json.loads(response.text)["body"])[
"landing_zone"
]
landing_zone = "contracts-landing-zone/"
except:
# st.write(myobj)
# st.write(response.text)
st.write("Internal Error. Reach out to Doczy.AI Team.")
batch_id = "failed_cases"
landing_zone = "contracts-landing-zone/"
else:
st.error("Failed")
for uploaded_file in file_list:
stringio = BytesIO(uploaded_file.getvalue())
stringio.seek(0)
s3_client.put_object(
Bucket=client_bucket,
Body=stringio.getvalue(),
Key=landing_zone + batch_id + "/" + str(uploaded_file.name),
)
s3_client.put_object_tagging(
Bucket=client_bucket,
Key=landing_zone + batch_id + "/" + str(uploaded_file.name),
Tagging={"TagSet": [{"Key": "BatchId", "Value": batch_id}]},
)
# TODO: Test this insert function with snowflake
upload_log = insert_upload_logs(
batch_id,
client,
str(uploaded_file.name),
datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
user_mail,
)
st.write(upload_log)
file_names.append(str(uploaded_file.name))
st.session_state.uploading = False
st.session_state.show_batchID = True
st.session_state.client_bucket = client_bucket
st.session_state.batch_id = batch_id
st.session_state.landing_zone = landing_zone
st.rerun()
if st.session_state.show_batchID:
batch_id = st.session_state.batch_id
client_bucket = st.session_state.client_bucket
landing_zone = st.session_state.landing_zone
st.write("Batch ID Created - ")
st.code(f"{batch_id}")
st.write(f"Files uploaded to s3://{client_bucket}/{landing_zone}{batch_id}")
st.session_state.show_batchID = False
st.session_state.file_list = []
st.session_state.batch_id = "failed_cases"
st.session_state.client_bucket = "default_bucket"
st.session_state.landing_zone = "contracts-landing-zone"
# RECOMMENDATION - Rethink about how to best manage cache
# @st.cache_data
# def convert_df(df):
# return df.to_csv(index=False).encode('utf-8')
# csv = convert_df(df)
# df = df.reset_index() # make sure indexes pair with number of rows
# contract_list = []
# for index, row in df.iterrows():
# group_list = []
# if row['Unique Key']:
# group_list.append('Unique Key')
# if row['Pricing Before Carveouts']:
# group_list.append('Pricing Before Carveouts')
# if row['Contract Related']:
# group_list.append('Contract Related')
# if row['Provider']:
# group_list.append('Provider')
# if row['Timeline']:
# group_list.append('Timeline')
# if row['Carveout Indicator']:
# group_list.append('Carveout Indicator')
# if row['Carveout Methodology']:
# group_list.append('Carveout Methodology')
# entry_dict = {
# "contract_name": row['Contract Name'],
# "groups": group_list,
# "contract_source_path": "batches/batch_1/"+client+"/"+row['Contract Name']
# }
# contract_list.append(entry_dict)
# contract_list = list(df['Contract Name'])
# myobj = {
# "s3_bucket": 'doczy-dev-infra-textract',
# "batch_id": "1",
# "client_name": client,
# "username": user_mail,
# "contract_list": contract_list
# }
# buttons = st.columns([0.8, 0.2])
# with buttons[0]:
# st.download_button("Download Table", csv, "file.csv", "text/csv", key='download-csv')
# with buttons[1]:
# if st.button("Upload to DB"):
# response = requests.post(doczy_pipeline, json = myobj)
# if response.status_code >= 200 and response.status_code < 300:
# st.write("Success")
# else:
# st.write("Failed")
# # st.write(response.text)