Merge branch 'FT-DEV-20/01/2026'

rag

Revert "refactor(ai-service): optimize prompts, token counting, and benchmarking - Externalize system prompts to text files for better maintainability. - Integrate tiktoken for faster and more accurate token counting. - Refactor BenchmarkLogger to use asynchronous logging for zero latency impact. - Improve prompt echo removal logic with more robust markers. - Add specialized system instruction for medical document processing. - Update test expectations to reflect precise token counting."

refactor(ai-service): optimize prompts, token counting, and benchmarking - Externalize system prompts to text files for better maintainability. - Integrate tiktoken for faster and more accurate token counting. - Refactor BenchmarkLogger to use asynchronous logging for zero latency impact. - Improve prompt echo removal logic with more robust markers. - Add specialized system instruction for medical document processing. - Update test expectations to reflect precise token counting.

class 'generator' error.

traceback in benchmark api

error success benchmark

exceptions and error logs added

bench mark

update unified manager.py

resolve streaming generation errors and improve robustness

dynamic cache fix

dynamic cache patch

openvino fix

changes for publishing the latest including generate_generic api

feat : SmartScribe HF AI API

use cache false

dynamic cache fixx

dynamic cache issue fix seen tokens

dynamic issue fix

dynamic cache issue fix

bypass casual openvino logic to load the models as transformer model

Add WSGI entrypoint and production Dockerfile for the AI service.

Implement generic text generation API with streaming support and add verification scripts

Add prompt template utility and unit tests for chat formatting

fix bear exceptions in unified Hugging Face Spaces deployment utilities, OpenVINO summarizer utils

Introduce AI medical extraction service with new API, agents, comprehensive testing, and documentation, while reorganizing existing scripts.

Enhance AI service by updating token limits for Phi-3.5 models, improving model loading logic, and refining prompt formatting in patient summary documentation.

feat: Add utilities for robust EHR data parsing, visit delta computation, and JSON serialization.

Enhance patient data processing by adding robust serialization and flattening utilities. The conversion function now captures all fields from EHR payloads, ensuring no data is lost during preprocessing. Additional logging improvements and error handling for plain text processing have been implemented.

Add robust JSON parsing utilities for medical data with flexible key matching and normalization capabilities.

Revert " Add robust JSON parsing utility for medical data with flexible key matching and data normalization."

Add robust JSON parsing utility for medical data with flexible key matching and data normalization.

Merge branch 'FT-DEV-26-11-2025'

feat: Add Colab patient summary script, AI service utilities for performance, and related documentation.

Merge branch 'FT-DEV-26-11-2025'

h immediate queue slot release and delayed job cleanup

Merge branch 'FT-DEV-26-11-2025'

feat: Add JobManager for asynchronous task tracking with auto-cleanup and a verification test.

add automatic cleanup for completed and error jobs, along with verification and demonstration tests.

Revert " implement `JobManager` for thread-safe asynchronous job tracking, including creation, updates, and cleanup, with supporting demo and test files."

Revert " implement `JobManager` for thread-safe asynchronous job tracking, including creation, updates, and cleanup, with supporting demo and test files."

implement `JobManager` for thread-safe asynchronous job tracking, including creation, updates, and cleanup, with supporting demo and test files.

Revert "feat: Introduce JobManager service for tracking asynchronous tasks with automatic cleanup and add dedicated tests."

feat: Introduce JobManager service for tracking asynchronous tasks with automatic cleanup and add dedicated tests.

Implement comprehensive AI model configuration, including T4 optimizations, token limits, and unified model management.

feat: Add prompt and token diagnostics to error responses for enhanced debugging.

feat: Implement unified model management, centralized constants, and error handling for AI medical extraction services.

1200 seconds from 600 secs timeout

Revert "feat: Establish AI medical extraction service with performance optimizations, unified model management, and detailed Hugging Face Spaces deployment guides."