How dharun11/ai works
This tree is a portfolio of AI applications rather than one unified runtime: resume review, student mentoring, document ingestion for retrieval, web research, and a SQL-oriented agent. The clearest user-to-result flows are resume text through parallel HR/engineering reviews and a merge, student chat through an agent orchestrator, uploaded documents through ingestion stages, and research sources through claim analysis to a brief. The README is only a heading; relationships below rely on the shown excerpts and SOURCE INDEX references. Some listed modules were not sampled, so uncertain wiring is omitted.
Main entry points: pipeline.py Ingestion pipeline · orchestrator.py Agent orchestrator · graph.py Research workflow · ingest.py Upload endpoint
Components 34
Resume Roaster
- Resume app
A2A/resume-roaster/app.py - Review orchestrator
A2A/resume-roaster/agents/orchestrator/main.py - Engineering review
A2A/resume-roaster/agents/engineer_agent/tools.py - HR review
A2A/resume-roaster/agents/hr_agent/tools.py
Document Ingestion
- Upload endpoint
Agentic Rag/src/api/routes/ingest.py - Pipeline composition
Agentic Rag/src/api/dependencies.py - Ingestion pipeline
Agentic Rag/src/ingestion/pipeline.py - Document parsing
Agentic Rag/src/ingestion/parsers/docling_parser.py - Chunk coordination
Agentic Rag/src/ingestion/chunkers/coordinator.py - Content deduplication
Agentic Rag/src/ingestion/hashing/mysql_store.py - Embedding providers
Agentic Rag/src/ingestion/embedders - Vector stores
Agentic Rag/src/ingestion/store - Ingestion records
Agentic Rag/src/ingestion/models.py - Run observability
Agentic Rag/src/ingestion/observability.py
Student Mentor
- Chat API
MentorAgent/main.py - Agent orchestrator
MentorAgent/orchestrator.py - Academic mentor
MentorAgent/agents/mentor_agent.py - Placement advisor
MentorAgent/agents/placement_agent.py - Emotional support
MentorAgent/agents/emotional_agent.py - Student data store
MentorAgent/db.py - LLM provider factory
MentorAgent/factory.py
Research Agent
- Research API
Research Agent/src/research_agent/api.py - Research workflow
Research Agent/src/research_agent/pipeline/graph.py - Source fetching
Research Agent/src/research_agent/fetch/fetcher.py - Claim extraction
Research Agent/src/research_agent/pipeline/extract.py - Facet gap analysis
Research Agent/src/research_agent/pipeline/gaps.py - Consensus synthesis
Research Agent/src/research_agent/pipeline/synthesize.py - Claim verification
Research Agent/src/research_agent/pipeline/verify.py
Other components
- Resume submitter
- Engineer A2A service
- HR A2A service
- Ingest client
- Student
- Research requester
Connections 33
- Resume submitter → Resume app: submits resume
- Resume app → Review orchestrator: invokes orchestrator
A2A/resume-roaster/app.py - Review orchestrator → Engineer A2A service: sends A2A review
A2A/resume-roaster/agents/orchestrator/main.py - Review orchestrator → HR A2A service: sends A2A review
A2A/resume-roaster/agents/orchestrator/main.py - Engineer A2A service → Engineering review: runs review
- HR A2A service → HR review: runs review
- Ingest client → Upload endpoint: uploads documents
- Upload endpoint → Pipeline composition: builds pipeline
Agentic Rag/src/api/routes/ingest.py - Upload endpoint → Ingestion pipeline: processes uploads
Agentic Rag/src/api/routes/ingest.py - Ingestion pipeline → Document parsing: parses documents
Agentic Rag/src/ingestion/pipeline.py - Ingestion pipeline → Chunk coordination: chunks content
Agentic Rag/src/ingestion/pipeline.py - Ingestion pipeline → Content deduplication: filters duplicates
Agentic Rag/src/ingestion/pipeline.py - Ingestion pipeline → Embedding providers: embeds chunks
Agentic Rag/src/ingestion/pipeline.py - Ingestion pipeline → Vector stores: upserts vectors
Agentic Rag/src/ingestion/pipeline.py - Upload endpoint → Ingestion records: formats results
Agentic Rag/src/api/routes/ingest.py - Upload endpoint → Run observability: counts outcomes
Agentic Rag/src/api/routes/ingest.py - Student → Chat API: asks question
- Chat API → Agent orchestrator: routes chat
MentorAgent/main.py - Agent orchestrator → Academic mentor: invokes agent
MentorAgent/orchestrator.py - Agent orchestrator → Placement advisor: invokes agent
MentorAgent/orchestrator.py - Agent orchestrator → Emotional support: invokes agent
MentorAgent/orchestrator.py - Agent orchestrator → Student data store: uses student data
MentorAgent/orchestrator.py - Agent orchestrator → LLM provider factory: creates model
MentorAgent/orchestrator.py - Academic mentor → Student data store: reads student records
MentorAgent/agents/mentor_agent.py - Placement advisor → Student data store: reads placement data
MentorAgent/agents/placement_agent.py - Emotional support → LLM provider factory: creates model
MentorAgent/agents/emotional_agent.py - Research requester → Research API: requests research
- Research API → Research workflow: starts workflow
- Research workflow → Source fetching: fetches sources
Research Agent/src/research_agent/pipeline/graph.py - Research workflow → Claim extraction: extracts claims
Research Agent/src/research_agent/pipeline/graph.py - Research workflow → Facet gap analysis: checks coverage
Research Agent/src/research_agent/pipeline/graph.py - Research workflow → Consensus synthesis: groups claims
Research Agent/src/research_agent/pipeline/graph.py - Research workflow → Claim verification: verifies claims
Research Agent/src/research_agent/pipeline/graph.py