Platform/Backend Engineer, GenAI Infrastructure

Lyric

full-time

Posted on: 9/22/2025

Origin: • 🇮🇳 India

✨ AI Apply

Mid-LevelSenior

CloudDistributed SystemsDockerGoGraphQLGRPCJavaKubernetesPythonScala

About the role

Design, build, and operate the backend services and platform APIs that power GenAI features in Lyric’s products.
Implement orchestration layers to route, sequence, and manage LLM calls with context grounding and prompt engineering.
Build scalable and cost-efficient GenAI serving infrastructure, including support for multiple model providers and fallback strategies.
Ensure platform resilience, security, observability, and compliance when serving user-facing GenAI workloads.
Provide abstractions, SDKs, and tooling to enable product and ML engineers to experiment and ship GenAI features faster.
Collaborate with ML engineers, product managers, and designers to understand requirements and deliver performant, developer-friendly systems.
Monitor performance, optimize latency and cost, and stay ahead of trends in GenAI and LLM ops.

3–9 years of backend or platform engineering experience, building distributed systems or ML/AI platforms.
Proficiency in at least one backend language (e.g., Python, Go, Java, Scala, or similar) and in designing APIs (REST, gRPC, GraphQL).
Experience building and operating scalable, high-availability backend services in cloud-native environments.
Familiarity with LLM serving and orchestration (e.g., OpenAI APIs, Anthropic, Hugging Face Inference Endpoints, or open-source LLM serving frameworks).
Understanding of prompt engineering, context grounding, and retrieval-augmented generation (RAG) concepts is a plus.
Knowledge of containerization and orchestration (Docker, Kubernetes) and observability best practices.
Ability to thrive in ambiguous, fast-moving environments and work effectively across teams.
Nice to have: Experience with vector databases (e.g., Pinecone, Weaviate, Milvus, FAISS) and embedding workflows.
Nice to have: Familiarity with LLM fine-tuning, adapters (LoRA, PEFT), or hosting custom models.
Nice to have: Knowledge of data privacy, security, and compliance concerns specific to GenAI workloads.