Gen AI Engineer with experience in building end-to-end document ingestion pipelines and RAG-based clinical knowledge assistants. Experienced in Python, Azure OpenAI Service, AWS Bedrock, FastAPI, REST APIs, Streamlit, Po…
Gen AI Engineer with experience in building end-to-end document ingestion pipelines and RAG-based clinical knowledge assistants. Experienced in Python, Azure OpenAI Service, AWS Bedrock, FastAPI, REST APIs, Streamlit, PostgreSQL, FAISS, ChromaDB, Pinecone, OpenSearch Vector DB, Talend, Power BI, Oracle, SQL Server, and related data integration tools.
<p>Worked closely with Business Analysts to review the business specifications of the project and also to gather the ETL requirements.<br>
Developed jobs, components and Joblets in Talend.<br>
Designed ETL Jobs/Packages using Talend Integration Suite (TIS).<br>
Used custom component tUnpivotRow.<br>
Used tStatsCatcher, tDie, and tLogRow to create a generic Joblet to store processing stats into a Database table to record job history.<br>
Created Talend Mappings to populate the data into dimensions and fact tables.<br>
Frequently used Talend Administrative Console (TAC).<br>
Extensively used the tMap component for lookup and joiner functions, along with tJava, tOracle, tXML, tDelimitedFiles, tLogRow, tLogBack, and other components across many jobs.</p>
<p>Built an end-to-end document ingestion pipeline using Python and LangChain for parsing, cleaning, and preprocessing enterprise documents.</p><p> Developed a RAG-based clinical knowledge assistant to help healthcare staff/analysts query medical policies, clinical guidelines, drug formularies, and patient care protocols.</p><p> Built ingestion pipelines to process structured and unstructured healthcare documents — clinical notes, discharge summaries, EHR extracts, and insurance policies.</p><p> Implemented document chunking, embeddings, and vector search using FAISS/Pinecone.</p><p> Integrated an LLM to generate contextual answers based on retrieved documents.</p><p> Designed prompts to reduce hallucinations and improve answer relevance.</p><p> Implemented source-document retrieval so users can validate generated responses.</p><p> Collaborated with clinical/compliance SMEs to validate chatbot responses against medical accuracy and regulatory standards.</p><p> Evaluated chatbot performance on healthcare-specific QA benchmarks, focusing on factual accuracy, safety, and reduction of clinically risky hallucinations.</p><p> Integrated domain-specific embeddings and retrieval strategies to improve accuracy on clinical and pharmacological queries.</p><p> Performed testing for retrieval accuracy, response relevance, hallucination, and edge cases.</p>
نظام التوظيف التشغيلي المعتمد على الذكاء الاصطناعي أولاً، الذي يوحّد ويؤتمت دورة التوظيف بالكامل للمؤسسات الحديثة حول العالم.
د.جوب منطقة حرة ذ.م.م. 2026 © جميع الحقوق محفوظة
