Required skills (Must-have)
Python & Engineering Fundamental
* sPython 3.12+, strong OOP, clean architecture
* .CLI scripting and orchestration (argparse, bash)
* .Package management, virtual environments, Docker
.
La descripción completa del puesto cubre todas las habilidades asociadas, la experiencia previa y cualquier cualificación que se espera que tengan los solicitantes.
Backe
* ndFastAPI (API design, service-oriented patterns
* ).Asynchronous processing: Celery, messaging, event-driven design (RabbitMQ / Azure Service Bus
* ).Relational DB experience, preferably PostgreSQ
L.
RAG / NLP /
* LLMRAG architecture and implementation (chunking, embeddings, retrieval, generatio
* n).Prompt engineering with versioning and iteration (multi-version prompt lifecycl
* e).OpenAI API experience including structured output parsi
* ng.Embeddings and retrieval tuning; hybrid retrieval approach (vector + BM2
5).
Vector & Se
* archQdrant (or Pinecone/Weaviate/Milvus) — collection management, metadata filt
* ers.Azure AI Search / Azure Cognitive Search (ACS) integration for indexing and retrie
val.
Document Proce
* ssingPDF extraction: PyMuPDF (f
* itz).OCR: Tesseract/pytesse
* ract.Chunking strategies optimized for legal/financial docs (clause/section aw
are).
Internal Tools
* / DemoStreamlit multi-page apps, session state, interactive grids (streamlit-ag
grid).
Data & Re
* portingpandas, NumPy, openpyxl — Excel report generation fro
* m JSON.Batch processing across multiple entities/pr
ojects.
Integration / Fullstack
* SupportBasic-to-intermediate Node.js and React (maintain/integrate existing se
rvices).
So
* ft skillsStrong communication and ability to work closely with accounting/audit SMEs to refine prompts and outputs (rapid iteration, POC envi
* ronment).Able to debug end-to-end pipelines independently (ingestion → retrieval → generation →
* output). xohynlm Strong analytical thinking and creative problem
-solving.