AI Backend Engineer

About the job

Experience: 4.00 + years Salary: INR 7000000-10000000 / year (based on experience) Expected Notice Period: 30 Days Shift: (GMT+05:30) Asia/Kolkata (IST) Opportunity Type: Hybrid () Placement Type: Full Time Permanent position (Payroll and Compliance to be managed by: Sedona.AI) (*Note: This is a requirement for one of Uplers' client - Sedona.AI) What do you need for this opportunity? Must have skills required: Python, Rust, Kubernetes, Terraform, Cloud, AI model deployment, Asr/tts, VLMs Sedona.AI is Looking for: We're looking for an AI Backend Engineer at Sedona AI to own the end-to-end inference infrastructure powering AI experiences for millions of homes. You'll design, build, and operate the systems that serve language, audio, and vision models in production - obsessing over p99 latency, cost efficiency, and reliability at scale. If you think deeply about distributed systems, model serving, and real-time data pipelines, and want to solve problems that don't have clean answers yet - this is the role. What You'll Work On • Design and own the backend systems that serve AI models in production — spanning language, audio, and vision-language models • Architect distributed serving infrastructure that handles real-world traffic reliably and efficiently • Build and optimize real-time audio processing pipelines for cloud-based inference • Serve vision-language and video understanding models at scale • Push the boundaries of inference performance — through smarter batching, caching, quantization, and serving strategies • Own the full stack from API design to model serving internals • Work directly with ML researchers and product engineers to take models from prototype to production What We're Looking For • 4+ years of experience in ML infrastructure, platform engineering, or large-scale backend systems • Strong foundations in systems engineering — distributed systems, backend design, performance optimization • Experience building and operating large-scale backend services in production • Hands-on experience deploying AI models using frameworks such as vLLM, SGLang, Triton, or equivalent • Real understanding of what it takes to serve models reliably — not just spin them up • Familiarity with real-time or streaming data processing, particularly for audio pipelines • Understanding of multimodal model architectures — audio (ASR, TTS), vision-language models (VLMs), and how they differ at inference time • Strong expertise in Kubernetes, Terraform, and cloud platforms (AWS and/or GCP) • Strong in Python and Rust; solid Linux fundamentals Nice to Have • Experience with inference optimization — KV-cache tuning, speculative decoding, dynamic batching • Familiarity with TensorRT-LLM or hardware-specific GPU optimizations • Experience with GPU cost modeling, multi-region inference, or traffic routing • Experience with high-availability system design or load balancing