Virender B.

10+ years of experience in architecting, designing and developing software across large scalable distributed systems and web applications. I am currently developing AI co-worker for Adobe, redesigning Agent Harness for long running always-on AI Agents. In my past experiences, I have been responsible for architecting AI Agent Harness, RAG Platform, MCP Registry for e-commerce giant in India, Flipkart. This harness is being used in 50+ agentic workflows within Flipkart Group. I have developed AI Avatar Solutions, AI Voice Agents Platform for realtime multi-lingual conversations. I specialise in Model Fine-tuning, model training from scratch, pre-traning, post-trianing, transfer learning , distillation techniques. Machine Learning: Neural networks, Deep learning, Transfer learning, Convolutional networks , Recurrent Neural Networks Gen AI: Conversation AI Agents, NVIDIA Omniverse, RAG, Model Context Protocol, auto-gen, Google-ADK Langchain, llama-index , A2A, Vector DBVector Store/Library : FAISS, Milvus, Qdrant, ChromaDB, Pinecone ML Enginerring : ML Training, ML Pipelines, Batch inference, NVIDIA, GPU, Time-slicing, Fractional-GPU Programming Languages : Python & Java Technologies: Apache Flink, Spark, Spring, Django, Flask Big Data : Flink, Spark, Hive, Hadoop, HBase Backend Engineering : FAST API, Django, Flask, Spring Boot Data/Databases : Druid, BigTable, BigQuery, MySQL, PostgreSQL, Hibernate Technologies : Kubernetes, Kafka, Redis, Elastic Search, RabbitMQ, Airflow, Azkaban Tools & Platforms : GCP, AWS, Azure, Terraform, Ansible, Prometheus, Grafana, Nginx, Gunicorn, Celery, JIRA, Git, CI/CD, TDD