diwasparashar
Compétences
Voir mes services

Expérience professionnelle
AI Engineer Intern
ZS • Temps plein
Mar 2026 - Present • 3 mos
Design, build, and deploy custom Python automation pipelines, predictive models, and Generative AI workflows for independent global clients. • Architected lightweight asynchronous backend servers and API systems using FastAPI and Python standard libraries, reducing client operational overhead. • Trained and optimized tree-based classification models (XGBoost, Random Forest) on unbalanced business datasets, utilizing SMOTE and threshold tuning to secure high classification benchmarks (up to 0.89 ROC-AUC). • Developed local Retrieval-Augmented Generation (RAG) systems and workflow automation templates using LangChain, vector databases (Qdrant, Chroma), and LLM APIs (Claude, Gemini). • Built automated document generation pipelines converting structured JSON outputs into print-ready, pixel-perfect formats, reducing manual document layout cycles.