Himansh Mudigonda

About Himansh Mudigonda

I’m a software engineer focused on backend systems and machine learning, with experience building practical products in fast-moving startup environments. My work has included backend services, ML-powered features, data pipelines, and developer-facing tools, with an emphasis on solving real problems and shipping reliable systems.I enjoy working across the stack when needed, but my strongest areas are Python backend development, applied machine learning, APIs, cloud infrastructure, and data-driven product engineering.I’m currently looking for Software Engineer, Backend Engineer, or Machine Learning Engineer roles where I can keep building, learning, and contributing. • : Python, C++, Java, Go, SQL. Currently: CUDA, Rust.• /: Agentic Workflows (LangGraph, MCPs), LLM Finetuning/Quantization, RAG, PyTorch, ONNX.• : AWS (EKS, CDK, SageMaker), GCP, Kubernetes, Docker, Terraform.• /: PostgreSQL, DynamoDB, Redis, VectorDBs (Pinecone, Chroma, Weaviate), Kafka, RabbitMQ

Experience

1. Software Engineer, Backend & Ml (Founding Team)

   Velociti

   Jun 2025 — Present · Phoenix, AZ, US

   Built the product’s core backend in Python, FastAPI and Rust, laying the foundation for pilot launch and helping the platform support 4x user growth during early adoption.• Developed and deployed reproducible AWS infrastructure using CDK, giving the team a version-controlled cloud environment and reducing manual configuration overhead.• Designed an asynchronous orchestration workflow using Kafka to support multi-step AI tasks more reliably, reducing timeout issues in longer-running request flows.• Improved model-serving performance by applying LLM quantization and inference optimizations, reducing P95 latency by 31% while also lowering GPU compute cost.• Built automated evaluation, centralized logging, and retry workflows for ML systems, helping catch regressions earlier and reducing MTTR by 80%.• Automated releases with GitHub Actions and AWS ECS, enabling the team to ship production updates more safely and consistently.

Education

• Lovely Professional University

  Bachelor of Technology - BTech