Data Science Engineer @Infor Decision Analytics and Science (IDeAS) at INFOR (2024-04 – Present)
Leading data science initiatives in pricing, forecasting, and inventory optimization for international clients across agriculture, retail, and manufacturing sectors.
- Designed and delivered end-to-end data-driven PoCs in pricing, forecasting, and inventory optimization for 5 international clients (agriculture, retail, manufacturing), aligning data science outcomes with business KPIs.
- Converted 2 PoCs into multi-year subscription deals: a potential $2M revenue uplift through pricing optimization (leveraging customer and product segmentation) for one of INFOR M3 ERP's top 10 global customers, and automated inventory policy generation for 4K+ item/warehouse intersections for a leading North American manufacturer.
- Led end-to-end delivery of 3 production releases for the subscribed North American manufacturer: two inventory optimization releases (v1.0 and v1.1 with enhancements) and a standalone sales forecasting use case, covering modeling through deployment.
- Combined regression-based (XGBoost) and time-series approaches (Prophet, Croston TSB) to address diverse demand patterns across datasets of up to 1–2M rows and 50–100K SKUs; managed model selection and validation via MLflow, Optuna, and demand-specific metrics (WAPE, BIAS).
- Engineered service-level constrained inventory optimization via Monte Carlo simulation over 100+ demand scenarios per SKU-location pair, identifying optimal (R,Q) and Min/Max policy parameters at scale.
- Engineered end-to-end batch ML pipelines on Infor OS Cloud Platform: ERP data ingestion, feature engineering, model training, and scheduled execution via ION orchestration; deployed models as REST API services; designed partition-level configurability allowing per-segment tuning across customer and product groups.
- Mentored junior engineers, conducted code reviews and technical interviews, contributed to architecture discussions, and onboarded multiple team members.
Software Data Engineer | Software Consultant @ALLIANZ & @CAHPP at ELYADATA (2022-09 – 2024-04)
Developed production-grade ML, NLP, OCR, and LLM microservices for insurance and healthcare clients.
- Built ML, NLP, OCR, and LLM microservices (regression, transcription, annotation, document parsing, GPT4All/Falcon inference) as production-grade REST APIs following a Backend for Frontend pattern.
- Built hybrid AI pipelines for healthcare (CAHPP) — combining NER, fuzzy matching, embeddings (pgvector), and LLMs for entity mapping across disparate sources, achieving >90% mapping accuracy.
- Supervised a final-year intern, conducted code reviews, and onboarded new team members.
Python Backend Engineer | NLP Data Scientist at UBIAI (2021-11 – 2022-09)
Architected and deployed NLP platform for document understanding with optimization focus.
- Architected and deployed an NLP platform for document understanding on AWS (EC2, Lambda, API Gateway), implementing NER, relation extraction, and text classification; optimized similarity computation by 27× through parallelization.
