Hi, I'm Ayush Saxena

Led the MLOps initiative for high-impact machine learning services, ensuring reliability and scalability in production environments.

• ML Pipeline Orchestration: Architected and maintained end-to-end ML pipelines utilizing Kubeflow and Airflow, reducing model deployment time by 60% and ensuring reproducibility.
• CI/CD & Automation: Designed robust CI/CD workflows using GitHub Actions and ArgoCD to automate testing, containerization (Docker), and deployment of ML models to Kubernetes clusters with rollback safety.
• Observability & Reliability: Implemented comprehensive monitoring using Prometheus, Grafana, and ELK stack; defined and monitored SLOs/SLIs to proactively manage error budgets and maintain 99.9% service availability.

Spearheading the engineering of SwissPlat, a next-gen B2B marketplace connecting suppliers with buyers through AI-driven matching and trade execution.

• Scalable Architecture: Developing a robust multi-tenant backend using Django and PostgreSQL, supporting isolated data environments for diverse enterprise clients.
• Advanced AI Features: Building RAG-based supplier matching and automated RFQ generation engines using OpenAI models, LangChain, and pgvector for semantic search.
• Performance & Security: Optimizing system throughput with asynchronous task queues (Celery/Redis) and securing the platform with enterprise-grade guardrails and prompt injection defenses.

Contributed to the development of large-scale data analytics platforms, enabling clients to derive actionable insights from terabytes of complex data.

• Cloud Infrastructure: Optimized data ingestion and processing workflows on AWS, utilizing EC2, S3, and Lambda, which resulted in a 30% reduction in processing latency.
• Security & Access Control: Strengthened cloud infrastructure security by implementing granular Role-Based Access Control (RBAC) and managing secure network configurations via Route53 and VPCs.