" I design and lead AI architectures that bridge technical rigor with societal impact. From healthcare diagnostics to financial reasoning platforms, I build scalable, observed, and ethical AI systems that empower organizations worldwide."

• Humanitarian Operations
• Healthcare & Medical Diagnostics
• Law & Access to Justice
• Marketing & Customer Insights
• Banking & Insurance
• Psychology & Sociology

• Description:Designed and delivered a dual-model AI diagnosis architecture on GCP, integrating real-time patient data ingestion, scalable inference via Ray Serve, and model lifecycle management with MLflow. Implemented high-availability failover using BentoML, containerized microservices, and CI/CD pipelines (Jenkins + Taskfile) for automated testing, deployment, and rollback. Ensured compliance with healthcare data regulations and optimized latency for clinical decision support.
• Tech Used: FastAPI · Ray Serve · MLflow · BentoML · Docker · Taskfile · Jenkins · Scikit-learn · Python
• Field : Healthcare

• Description: Designed and implemented the Data Engineering backbone of an enterprise-grade Finance RAG architecture, delivering robust ingestion, chunking, and embedding pipelines with distributed orchestration and full observability. This foundation ensures reliable data flow and scalability for the upcoming retrieval and LLM reasoning layers within a modular GenAI ecosystem.
• Tech Used:Python · RabbitMQ · Docker · MinIO · Prometheus · Grafana · Elasticsearch · Kibana · Fluentd · LangChain · Milvus · SpaCy · PyTorch · TensorFlow · Taskfile · CI/CD (Jenkins)
• Field : Finance

• Description: Designed a cloud-native churn prediction architecture integrating batch analytics and real-time API inference for e-commerce customer segmentation. Implemented an API-first approach with FastAPI services, orchestrated CI/CD (Jenkins) for both backend and Django-based frontend, and a model failover strategy using BentoML. Monitored model performance and operational health with Prometheus and Grafana, ensuring 97% precision sustained post-deployment.
• Tech Used: Scikit-learn · FastAPI · Django · Docker · MLflow · BentoML · Prometheus · Grafana · CI/CD (Jenkins) · Python
• Field : Marketing - E-commece sector

• Event-driven microservices with Kafka & RabbitMQ for real-time ingestion
• RAG (Retrieval-Augmented Generation) pipelines for finance and knowledge management
• Hybrid batch + streaming architectures for analytics and real-time APIs
• Cloud-native ML deployment with BentoML + Kubernetes
• Observability-first design: Prometheus, Grafana, OpenTelemetry integrated from the start
• Failover & resilience patterns: multi-model fallback, autoscaling, CI/CD rollback strategies

• Mentored junior engineers and data scientists in building ML pipelines
• Led internal workshops on MLOps, CI/CD, and observability practices
• Contributed to DevOps/AI culture through documentation, code reviews, and best practices
• Coordinated cross-functional teams (IT, marketing, healthcare, finance) to align tech with business goals
• Established architecture standards ensuring scalability and consistency across projects

• 97% precision sustained in production for churn prediction models
• 40% latency reduction on medical diagnosis APIs (Ray Serve + FastAPI optimization)
• 1M+ events/day processed through Kafka + Elasticsearch pipelines
• 20% cloud cost savings achieved via orchestration and proactive monitoring
• 99.9% uptime maintained on critical microservices with autoscaling and failover
• <5 min rollback guaranteed through CI/CD pipelines (Jenkins + Taskfile)