AVANEESH SINGH

Led a 6-developer team in redesigning and delivering a high-volume Supplier Feed batch ingestion pipeline, leveraging event-driven architecture with Azure Blob Storage and Functions to process up to 1000 uploads/day.

Achieved 99.999% inter-service consistency and reliable post-commit event delivery by implementing the Outbox Pattern across critical microservices.

Spearheaded cross-team database optimization initiatives, enhancing API responsiveness by 48% across UK, EU, and JP regions through advanced query plan analysis and indexing strategies.

Drove backend maintainability initiatives by resolving critical Sonar issues and increasing test coverage to 90%, enhancing code quality and reducing technical debt.

Directed a 4-member team in developing and designing high-performance, scalable RESTful APIs utilizing C# .NET and CQRS for a pharmaceutical client.

Engineered a C#-based automated ETL conversion from Informatica to Talend, reducing manual effort and streamlining data integration workflows by 95% and improving project operational efficiency by over 50%.

Streamlined secure user authentication by integrating OAuth 2.0 Device Flow with Keycloak, ensuring compliance and robust access control.

Mentored and accelerated junior engineer proficiency by conducting 30+ hours of technical training on architectural design and coding standards, fostering team growth and project expertise.

Developed and benchmarked a Proof-of-Concept (PoC) data pipeline, integrating NATS and Kafka within a microservice architecture to rigorously assess scalability, throughput, and latency across critical services.

Optimized messaging system middleware for low-latency streams and enhanced throughput comparison, significantly improving data processing efficiency.

Provided actionable insights and strategic recommendations derived from PoC benchmarks, directly influencing decision-making for production-grade system implementations.

Developed advanced deep learning models (VGG16, DenseNet169) for facial expression classification, achieving up to 85% accuracy in complex scenarios.

Enhanced model transparency and interpretability by implementing explainability techniques (Grad-CAM, feature visualization) to articulate neural network decision-making processes.