o ;Qh @sddlmZddlmZddlmZddlmZddlm Z ddl Z ddl Z ddl m Z mZmZddlmZeGd d d ZGd d d eZGd ddeZGdddZddZddZddZedkr ededdZedededdZdd d!d"d#d!d$d%d!d&d'd!d(d)d!d*d+d!d,d-d!gZd.Zd/Zed0ed1eed2eed3eed4eej eeeed5Z!d+Z"ej#ee"eed6Z$ed7ee$ed8d9d:e$d;DZ%ed<ee%d=d>d?d@dAdBdCdDdEdFdG Z&eee%e&Z'edHdSdS)I)OllamaPromptTemplate)BaseOutputParser)LLMChain) ChatOllamaN)DictListOptional) dataclassc@s&eZdZUeed<eed<eed<dS) SkillDatanamecountlevelN)__name__ __module__ __qualname__str__annotations__intrrH/var/www/eduai.edurigo.com/doc_train/edurigo_ai/TNA/testing/tna_final.pyr s  r c@4eZdZdZdedefddZdedefddZdS) DepartmentOutputParserz-Parser for Step 1 - Department Identificationtextreturnc Cslztd|tj}|rt|WS||WSty5}zdddt|dWYd}~Sd}~ww)N\{.*?\}Unknownlowz Parse error: identified_department confidence reasoning) researchDOTALLjsonloadsgroup_extract_department_from_text Exceptionr)selfr json_matcherrrparses  zDepartmentOutputParser.parsecCs ddddS)Nrmediumz#Could not parse structured responserr)r+rrrrr)#sz4DepartmentOutputParser._extract_department_from_textN)rrr__doc__rrr.r)rrrrrsrc@r) SkillAnalysisOutputParserz+Parser for Step 2 - Skill Priority Analysisrrc Csz?td|tj}|rt|}d|dWStd|tj}|r:t|}dt|tr5|gdWS|dWS||WSt yY}zdt |gdWYd}~Sd}~ww)Nz\[.*?\]T)successskill_recommendationsrFr2errorr3) r#r$r%r&r'r( isinstancedict_extract_skills_from_textr*r)r+rr, skills_data skill_datar-rrrr.-s. zSkillAnalysisOutputParser.parsec sg}|d}i}|D]rdrq td}|r8|r.d|vr.||d|di}q tfdddDrc|rPd|vrP||i}td d d d }||d<q d  vr{t d}|rzt |d|d<q tfdddDrd d tfdddDr|d<q d vrt d}|rt |d|d<q |rd|vr|||D]&} | dd| dd| dd| dd| dd| ddq|rd nd!|t |d"S)#z,Enhanced text extraction for multiple skills #z ^\d+\.\s*(.+) skill_namec3|] }|vVqdSNlower.0keywordlinerr `zFSkillAnalysisOutputParser._extract_skills_from_text..)zskill:z skill_name:- z^[\-\*\d\.\s]*:priorityz(\d+(?:\.\d+)?)priority_percentagec3r?r@rArCrFrrrHmrI)zlevel:requiredc3r?r@rA)rDr) level_textrrrHorI)beginner intermediateexpertrequired_levelcoveragetarget_coveragegI@ Intermediatecurrent_coverageggN@gap_typez Missing Skillr"z)Standard skill requirement for departmentTF)r2r3extracted_count)stripsplit startswithr#matchappendr(anysubrBr$float setdefaultlen) r+rskillslines current_skillnumbered_matchr=priority_matchcoverage_matchskillr)rQrGrr8Js^                  z3SkillAnalysisOutputParser._extract_skills_from_textN)rrrr0rrr.r8rrrrr1*sr1c @sHeZdZd-dedefddZdefdd Zdefd d Zd ed ee de fddZ dedee de e fddZ dede deede fddZd ed ee de dede f ddZd edede dede f ddZdeede defd d!Zdedeefd"d#Zd$edefd%d&Zd'ee de dee fd(d)Zd'ee de fd*d+Zd,S).TwoStepTNASystem gemma3:12bhttp://localhost:11434 model_namebase_urlcCsf||_||_t||ddd|_t|_t|_t|j| |jdd|_ t|j| |jdd|_ dS)z Initialize Two-Step TNA System Args: model_name: Ollama model name base_url: Ollama server URL g?i)modelrq temperature num_predictF)llmprompt output_parserverboseN) rprqrrur dept_parserr1 skill_parserr_create_department_prompt dept_chain_create_skill_analysis_prompt skill_chain)r+rprqrrr__init__s, zTwoStepTNASystem.__init__rcCsd}tddg|dS)z=Create prompt template for Step 1 - Department Identificationas You are an expert at identifying organizational departments based on user descriptions. TASK: Analyze the user prompt and identify which specific department they are referring to from the organization's department list. ORGANIZATION DEPARTMENTS: {organization_departments} USER PROMPT: "{user_prompt}" INSTRUCTIONS: 1. Carefully analyze keywords, responsibilities, and context in the user prompt 2. Match the description to the most appropriate department from the provided list 3. Consider typical department functions and responsibilities 4. Choose the best match from the given department list 5. Provide confidence level and clear reasoning RESPOND IN THIS EXACT JSON FORMAT: {{ "identified_department": "exact department name from the list", "confidence": "high/medium/low", "reasoning": "brief explanation of why this department was chosen" }} JSON Response: user_promptorganization_departmentsinput_variablestemplaterr+rrrrr{s z*TwoStepTNASystem._create_department_promptcCsd}tgd|dS)z;Create prompt template for Step 2 - Skill Priority Analysisa You are an expert Training Need Analysis (TNA) consultant specializing in skill gap analysis. TASK: Analyze the current skills for the {department} department and recommend skill priorities that need to be addressed. DEPARTMENT: {department} TOTAL EMPLOYEES: {employee_count} CURRENT SKILLS: {current_skills} ANALYSIS REQUIREMENTS: 1. COVERAGE ANALYSIS: - Calculate current skill coverage percentages - Identify skills with insufficient coverage - Standard target: 50-70% coverage for most skills, 70%+ for critical skills 2. LEVEL ANALYSIS: - Assess if current skill levels are appropriate for the department - Consider if Beginner/Intermediate/Expert levels match job requirements 3. MISSING SKILLS: - Identify essential skills that are completely missing - Consider both technical and soft skills needed for this department 4. PRIORITY CALCULATION: - Base priority on: coverage gaps, level gaps, business impact, urgency - Priority scale: 90-100% (Critical), 70-89% (Important), 50-69% (Moderate), 30-49% (Nice-to-have), 10-29% (Low) 5. DEPARTMENT-SPECIFIC FOCUS: - Consider the core functions and responsibilities of {department} - Include industry-standard skills for this department type - Think about both current needs and future requirements IMPORTANT: You MUST provide EXACTLY 10-15 skill recommendations. Include: - Analysis of existing skills that need improvement - Essential missing skills for this department - Both technical and soft skills - Skills at different priority levels RESPOND WITH JSON ARRAY OF EXACTLY 10-15 SKILL RECOMMENDATIONS: [ {{ "skill_name": "exact skill name", "required_level": "Beginner/Intermediate/Expert", "priority_percentage": 85.5, "current_coverage": 25.0, "target_coverage": 70.0, "gap_type": "Coverage Gap/Level Gap/Missing Skill/Enhancement", "reasoning": "detailed explanation of why this skill needs attention and the assigned priority" }}, {{ "skill_name": "another skill name", "required_level": "Intermediate", "priority_percentage": 78.0, "current_coverage": 0.0, "target_coverage": 60.0, "gap_type": "Missing Skill", "reasoning": "explanation for this skill priority" }} ] CRITICAL: Provide MINIMUM 10 skills, MAXIMUM 15 skills. Do not provide fewer than 10 recommendations. Sort by priority percentage (highest first). Include a variety of skills: existing skills needing improvement + completely missing essential skills. JSON Response: departmentemployee_countcurrent_skillsrrrrrrr}s Dz.TwoStepTNASystem._create_skill_analysis_promptrrc CsPztdtd|tdtdd|Ddd|D}|d|d}|j|}t|tr|d d }|||}|rn|d |d <|d |d <||d<d|d<td|d d|d d|ddd|WSd|d<d|d|d<d|d <d|d <|WSdddddWSt y}zdt|dddWYd}~Sd}~ww)a STEP 1: Use LLM to identify department from user prompt Args: user_prompt: User's description/prompt about the department organization_departments: List of department dictionaries with 'id' and 'name' keys Example: [{'id': 1, 'name': 'HR'}, {'id': 2, 'name': 'Engineering'}] Returns: Dict with identified department info, confidence, and reasoning u2🔄 STEP 1: Identifying department from prompt...z Prompt: z Available Departments: cSs$g|]}|dd|ddqS)r  (ID: id)rrDdeptrrr /s$z>TwoStepTNASystem.step1_identify_department..cSg|]}|dqS)r rrrrrr5, )rrr rKr r department_iddepartment_infoTr2u!✅ STEP 1 Complete: Identified 'z' (ID: z) with r!unknownz confidenceFz Department 'z ' not found in organization listr5Nz2Failed to parse department identification response)r2r5r r) printrjoinr|runr6r7get_match_department_to_listr*) r+rrdepartment_names chain_inputresultidentified_dept_namematched_dept_infor-rrrstep1_identify_department sJ       ,z*TwoStepTNASystem.step1_identify_departmentidentified_dept dept_listcCsp|sdS|}|D]}|d|kr|Sq |D]}|d}||vs1||vr5|SqdS)aU Match identified department with organization's department list Args: identified_dept: Department name identified by LLM dept_list: List of department dictionaries with 'id' and 'name' keys Returns: Dict with department info if found, None otherwise Nr )rBr\)r+rridentified_lowerrdept_name_lowerrrrrbs  z*TwoStepTNASystem._match_department_to_listrrrc Csztd|dtd|tdt|d|||}|||d}|j|}tdt|tdt|d d d t|tr| d rt| d g}td|d| |d |}d||t||| |t|j |t|dd} tdt|d| WStdt|} |j | } | d r| d r| | d |}tdt|dd||t||| |d| ddddWSddgdtt|dkrt|d dd nt|idWSty} zdt| gd WYd } ~ Sd } ~ ww)!a STEP 2: Use LLM to analyze skills for the identified department Args: department: Department identified from Step 1 employee_count: Total number of employees in department current_skills: List of current skills with counts and levels Returns: Dict with skill priority recommendations u"🔄 STEP 2: Analyzing skills for z department... Employee Count:  Current Skills: z skills providedru 🔍 Debug - LLM Response type: u🔍 Debug - Response preview: Nz...r2r3u🔍 Debug - Parsed z skills from JSON responseT)raw_response_typeraw_skill_countprocessed_skill_count)r2rrtotal_recommendationsr3analysis_summary debug_infou✅ STEP 2 Complete: Generated z skill recommendationsuY⚠️ Warning: Received non-dict or unsuccessful response, attempting manual parsing...u0✅ STEP 2 Complete (manual parsing): Generated manual_extractionr[r)parsing_methodr[FzIFailed to parse skill analysis response with both JSON and manual methods raw_responsei)r2r5r3rr4)rre_format_skills_for_llmr~rtyperr6r7r_post_process_recommendations_generate_summaryrrzr8r*) r+rrrskills_descriptionrr skill_countprocessed_recs final_resultraw_text manual_resultr-rrrstep2_analyze_skillss|     , z%TwoStepTNASystem.step2_analyze_skills skills_inputc Cstdtd|||}|dsdd|dd|dS|d }|d }|d }td |d |dtd||} |||| } | dsVdd| dd|| dSd|| | d| ddS)a" Complete two-step TNA analysis Args: user_prompt: User's description of department/team organization_departments: List of department dictionaries with 'id' and 'name' keys Example: [{'id': 1, 'name': 'HR'}, {'id': 2, 'name': 'Engineering'}] employee_count: Total employees in the department skills_input: Current skills data (formatted string) Returns: Complete analysis results from both steps u&🚀 Starting Complete TNA Analysis...zP================================================================================r2Fr>r5z Department identification failed)r2stepr5 step1_resultr rru✅ Department Identified: rrSkill analysis failed)r2rr5r step2_resultTr3r)r2rrfinal_recommendationsr)rrr_parse_skill_inputr) r+rrrrrr rrrrrrrcomplete_tna_analysiss:       z&TwoStepTNASystem.complete_tna_analysisr cCsN||}||||}|dsdd|dd|dSd||d|d d S) Nr2Frr5r)r2rr5rTr3r)r2rrr)rrr)r+rr rrrrrrrtna_skill_analysis_use"s    z'TwoStepTNASystem.tna_skill_analysis_userfcCs`|sdSg}|D]"}|j|d}|d|jd|jd|d|dd|jd qd |S) z#Format current skills for LLM inputz No current skills data provided.drJz: /z employees (z.1fz% coverage) at z levelr;)rr`r rr)r+rfr formattedrlrVrrrr9s* z'TwoStepTNASystem._format_skills_for_llmc Csg}|s|Std|}|D]-}td|}|r?|d}t|d}||d}|t|||q|S)z/Parse skill input string into SkillData objectsz,\s*z(.+?)-(\d+)\s*\((.+?)\)r>r) r\r#r]r_r(r_normalize_skill_levelr`r ) r+rrf skill_partspartr_r=rrrrrrGsz#TwoStepTNASystem._parse_skill_inputrcCsH|}|dr dS|dsd|vrdS|ds |dr"dSdS) z(Normalize skill level to standard formatbeginBeginnerinterr/rXrTadvExpert)rBr^)r+r level_lowerrrrrZs z'TwoStepTNASystem._normalize_skill_levelrecommendationscCsg}|D]g}zZ|dd||ddtdtdt|ddtdtdt|d dtdtdt|d d|d d|d d d}t|d d||d<||Wqtt fykYqw|j dddd|S)z)Post-process and validate recommendationsr=rrUrXrrrO2rYrWrZr"zNo reasoning provided)r=rUrOrYrWrZr"target_employee_countcSs|dS)NrOr)xrrrsz@TwoStepTNASystem._post_process_recommendations..T)keyreverse) rr\rmaxminrcrr` ValueError TypeErrorsort)r+rr processedrec processed_recrrrres( z.TwoStepTNASystem._post_process_recommendationscCs|sddiStdd|D}tdd|D}tdd|D}tdd|Dt|}t||||d t|d |rD|d d dSd dS)zGenerate analysis summarymessagezNo recommendations generatedcSsg|] }|ddkr|qS)rOPrrDrrrrrz6TwoStepTNASystem._generate_summary..cSs,g|]}d|dkrdkrnn|qS)rrOrrrrrrrs,cSsg|] }|ddkr|qS)rOrrrrrrrrcss|]}|dVqdS)rONrrrrrrHsz5TwoStepTNASystem._generate_summary..) high_prioritymedium_priority low_priorityr>rr=N) total_skillspriority_distributionaverage_prioritytop_priority_skill)resumround)r+rrrr avg_priorityrrrrs z"TwoStepTNASystem._generate_summaryN)rnro)rrrrrrr{r}r rrr rrr rrrrrrrrrrrrrms<'!LB  f =  rmcCsn|r|stdt|}|d}|d}tddgdd}tgdd d}td d d }|} ||B} | |d } ||B} | ||d} | | }| | }|j}|j}|dd}|dd}t|}t dt |t dt|}| dg}| dd}| di}| dd}| dg}|p|}dd|D}d|||||di}t tj |dd|S)N0Goal/problem and missing skills must be providedformatted_stringskills_with_levelsgoalmissing_skills_stra, You are a professional learning and development advisor. Given the user's business goal, and their identified missing skills WITH REQUIRED LEVELS, output the recommendation report in **STRICT JSON FORMAT**. ### JSON Output Format: {{ "Goal":{{ "goalTitle": "string", "course_recommendations_online": [ {{ "platform": "Platform name", "course_name": "Course title", "level": "Beginner / Intermediate / Advanced", "target_skill": "Skill name being addressed", "required_level": "Required proficiency level for this skill", "why_recommended": "Reason why the course is useful for achieving the required skill level", "relevance_score": integer (0-100), "link": "Course URL or 'Available on [Platform]'" }} ] }} }} **Rules:** - Only return valid JSON, no markdown, no headings. - Use double quotes `"` around all keys and string values. - Ensure the JSON is syntactically correct (ready for `json.loads()`). - If a field is missing, use an empty string "" or empty array []. - Only give those courses which relevance_score more than 75 - Consider the required proficiency level when recommending courses - Match course difficulty to the required skill level --- ### Inputs: Goal / Problem: {goal} Missing Skills with Required Levels: {missing_skills_str} rrrcourse_data_jsona You are a professional learning and development advisor. You are given: - A business goal - A list of missing skills WITH REQUIRED LEVELS - A database of available courses (with name and description) ### Business Goal: {goal} ### Missing Skills with Required Levels: {missing_skills_str} ### Available Courses: {course_data_json} Your task is to: 1. Evaluate each course for how well it helps achieve the goal and address the missing skills at the required proficiency levels. 2. Select the most relevant courses (maximum 5). 3. For each course, return ONLY the following fields: - "course_name": string - "course_id": string - "target_skill": string (which skill this course addresses) - "required_level": string (required proficiency level for the skill) - "reason_to_recommend": string - "relevance_score": integer (0-100) Don't give any extra text. Just return JSON Output. ### JSON Output Format: {{ "Goal":{{ "goaltitle": "string", "course_recommendations_offline": [ {{ "course_name": "Course title", "course_id": "Course ID from course_data_json", "target_skill": "Skill name being addressed", "required_level": "Required proficiency level for this skill", "why_recommended": "Reason why the course is useful for achieving the required skill level", "relevance_score": integer (0-100), "link": "Course URL or 'Available on [Platform]'" }} ], "suggested_order": [ "Course 1 name", "Course 2 name", "Course 3 name" ] }} }} **Rules:** - Only return valid JSON, no markdown, no headings. - Use double quotes `"` around all keys and string values. - Ensure the JSON is syntactically correct (ready for `json.loads()`). - If a field is missing, use an empty string "" or empty array []. - Only give those courses which relevance_score more than 75 - Consider the required proficiency level when recommending courses rnrrrrsrr```json```z`=============================================================================Offline Goal Start z]=============================================================================Offline Goal Endcourse_recommendations_offline goaltitlerKGoal goalTitlecourse_recommendations_onlinecSr)r=r)rD skill_inforrrrLrz=recommend_online_offline_courses_goal_try..)rmissing_skillsmissing_skills_with_levelsrrrindent) r"process_missing_skills_with_levelsrrinvokecontentr\r&r'rrdumps)rrcourses_availableprocessed_skills_inforrcourse_recommendation_prompt2'course_recommendation_prompt_simplifiedrugoal_stronline_course_extraction_chaininput_data_onlineoffline_course_extraction_chaininput_data_offlinerecommendations_offlinerecommendations_onlinerecommendations_offline_data1recommendations_online_data1recommendations_offline_data2recommendations_online_data2recommendations_offline_datarecommendations_online_datargoal_title_offlinegoal_data_onlinegoal_title_onlinerfinal_goal_titlemissing_skills_list final_outputrrr)recommend_online_offline_courses_goal_trysn0B          r!cCsg}t|trdd|dD}dd|D}n8t|trQ|D]0}t|tr0||ddq t|trF||dd|d ddq |t|ddq g}|D]}||dd |d d qUd |}||d S)z Process missing skills input to handle different formats: - List of skill names: ["CRM Tools", "Communication"] - List of skill objects: [{"skill_name": "CRM Tools", "required_level": "Intermediate"}] - Mixed format cSg|]}|qSrr\rDrlrrrrhrz6process_missing_skills_with_levels..,cSsg|]}|ddqS)rXr=rUr)rDr rrrrisrXr&r=rKrUz (Required Level: rr)rr)r6rr]listr`r7rr)rr skill_namesrlformatted_skillsrrrrrr]s6          rcCs|r|stdtddgdd}tgddd}tdd d }|}|}||B}||d } ||B} |||d} | | } || } | j}| j}|d d }|d d }t|}t|}|dg}|dd}|di}|dd}|dg}|p|}t |t rdd| dD}nt |t rt dd|Drdd|D}n|}tdt|tdt||||d}ttj|dd|S)Nrrra You are a professional learning and development advisor. Given the user's business goal , and their identified missing skills, output the recommendation report in **STRICT JSON FORMAT**. ### JSON Output Format: {{ Goal:{{ "goalTitle": "string", "course_recommendations_online": [ {{ "platform": "Platform name", "course_name": "Course title", "level": "Beginner / Intermediate / Advanced", "why_recommended": "Reason why the course is useful", "relevance_score": integer (0-100), "link": "Course URL or 'Available on [Platform]'" }} ] }} }} **Rules:** - Only return valid JSON, no markdown, no headings. - Use double quotes `"` around all keys and string values. - Ensure the JSON is syntactically correct (ready for `json.loads()`). - If a field is missing, use an empty string "" or empty array []. - Only give those courses which relevance_score more than 75 --- ### Inputs: Goal / Problem: {goal} Missing Skills: {missing_skills_str} rrag You are a professional learning and development advisor. You are given: - A business goal - A list of missing skills - A database of available courses (with name and description) ### Business Goal: {goal} ### Missing Skills: {missing_skills_str} ### Available Courses: {course_data_json} Your task is to: 1. Evaluate each course for how well it helps achieve the goal and address the missing skills. 2. Select the most relevant courses (maximum 5). 3. For each course, return ONLY the following fields: - "course_name": string - "course_id" : string - "reason_to_recommend": string - "relevance_score": integer (0-100) Don't give any extra text . Just return JSON Output. Here is the JSON class to validate the output that you give. Make sure the check always passes. class SimpleOfflineCourseRecommendation(BaseModel): ### JSON Output Format: Goal:{{ "goaltitle": "string", "course_recommendations_offline": [ {{ "course_name": "Course title", "course_id" : Course ID from course_data_json, "why_recommended": "Reason why the course is useful", "relevance_score": integer (0-100), "link": "Course URL or 'Available on [Platform]'" }} ], "suggested_order": [ "Course 1 name", "Course 2 name", "Course 3 name" ] }} **Rules:** - Only return valid JSON, no markdown, no headings. - Use double quotes `"` around all keys and string values. - Ensure the JSON is syntactically correct (ready for `json.loads()`). - If a field is missing, use an empty string "" or empty array []. - Only give those courses which relevance_score more than 75 rnrrrrrrrrKrrrcSr"rr#r$rrrr<rz9recommend_online_offline_courses_goal..r%css|]}t|tVqdSr@)r6r')rDitemrrrrH=sz8recommend_online_offline_courses_goal..cSsg|] }|D]}|qqSrr)rDsublistrlrrrr?s!course_recommendations_offline234 course_recommendations_online234)rrrrr)rrrrrr\r&r'rr6rr]r'allrr )rrr r r rurrrrrrrrrrrrrrrrrrrrrr rrr%recommend_online_offline_courses_goalsl-C          r/__main__z#Initializing Two-Step TNA System...rn)rpze ====================================================================================================z'EXAMPLE: COMPLETE TWO-STEP TNA ANALYSISzd====================================================================================================zZWe need training for our team that handles customer complaints and manages support ticketsr>Sales)rr r MarketingrzHuman ResourceszInformation TechnologyFinancezCustomer Service OperationszxCommunication-12(Intermediate), Active Listening-6(Intermediate), CRM Tools-5(Beginner), Problem Solving-8(Intermediate)u📝 Input Data:z User Prompt: z Organization Departments: rr)rrrr)rr rrz8--------=-=-=-=-=-=---------------------================zO++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++=cCs$g|]}|d|d|ddqS)r=rUr")r=rUr"rr$rrrrsrrrzoLearn the fundamentals of Python programming, including variables, loops, functions, and basic data structures.zqCovers the basics of data analysis, statistics, and visualization using tools like Pandas, NumPy, and Matplotlib.zhBuild and evaluate models using regression, classification, and clustering techniques with Scikit-Learn.zWExplore neural networks, CNNs, and RNNs using TensorFlow to build deep learning models.zgUnderstand text preprocessing, embeddings, and LLMs using libraries like NLTK, spaCy, and Hugging Face.zpMaster SQL queries, joins, subqueries, and window functions for effective data analysis on relational databases.z`Create and deploy AI-driven RESTful APIs using FastAPI and integrate them into web applications.z]Learn to build Retrieval-Augmented Generation pipelines using LangChain, FAISS, and ChromaDB.zaUnderstand core Azure services, AI tools, and deployment techniques for scalable AI applications.zeLearn graph theory concepts and build intelligent applications using Neo4j and Cypher query language.) zPython for BeginnerszData Science Essentialsz"Machine Learning with Scikit-LearnzDeep Learning with TensorFlowzNatural Language ProcessingzSQL for Data Analysisz AI-Powered Web Apps with FastAPIzLangChain and RAG Pipelinesz Cloud Fundamentals with Azure AIzGraph Databases with Neo4jcourses_online_offline)(langchain_community.llmsrlangchain.promptsrlangchain.schemarlangchain.chainsrlangchain_ollamarr&r#typingrr r dataclassesr r rr1rmr!rr/rr tna_systemrrrrrrr rresult1rcourse_availabler:rrrrs     aE/ L