Senior Software Engineer with 8+ years of experience specializing in backend engineering, AI, cloud engineering, and distributed systems. I currently work at CodiuX, where I lead the architecture and product strategy for AI-powered voice platforms. My experience spans SaaS, Voice AI, IoT, cloud platforms, and enterprise systems.

I design multi-agent and distributed architectures, real-time conversational AI systems, microservices, and scalable cloud infrastructure — and enjoy collaborating with engineering teams to build reliable, production-grade software. My background includes leading backend teams, developing cloud-native platforms, and supporting cross-functional product delivery.

Languages Python · JavaScript/TypeScript · Rust · Go · SQL

Backend & APIs Django · Django REST Framework · FastAPI · Flask · Node.js · Express.js · NestJS · Gin · Spring Boot

Rust Tokio · Axum · Actix-web · Tonic · Hyper · Tower · Serde · sqlx · nom · prost cbindgen · bindgen · Flume · Crossbeam · Rayon · Criterion · Proptest · cargo-nextest

Frontend React.js · Next.js · Angular

AI & Voice LLMs · RAG · LangChain · LangGraph · Voice AI (LiveKit) · Speech Processing · NLP Multi-LLM agent orchestration · Anthropic Claude · OpenAI · tool calling · agent evaluation harnesses

Data & Messaging PostgreSQL · MongoDB · Redis · Kafka · GraphQL · gRPC · WebSockets

Cloud & Infrastructure AWS (Lambda, SQS, S3, RDS, ECS) · GCP (GKE) · Kubernetes · Docker · Terraform · Datadog

Systems SIP/WebSocket gateways · binary-frame parsers · FFI bridges · zero-copy audio pipelines · async event buses · streaming inference runtimes

Senior Software Engineer at CodiuX, leading architecture and product strategy for AI-powered voice platforms — production voice AI systems handling 2,000–5,000 concurrent sessions/hr at p99 < 40 ms, SIP/WebSocket gateways, multi-LLM agent orchestration, and HIPAA/GDPR compliance.

QuicMQ — QUIC message broker giving each subscription its own stream to avoid the head-of-line blocking Kafka and NATS have. WAL durability, replay, Prometheus metrics, and a CLI.

sesame-csm — Streaming inference runtime converting batch-only voice AI models into sub-second backends. Tokio pipeline, zero-copy audio buffers, SIMD frame processing, Tonic gRPC interface.

linkedin.com/in/awaiskhan404 · awais.push@gmail.com Multan, Pakistan — open to remote / relocation Open to senior engineering roles in AI backend, cloud, and distributed systems.