Backend AI Engineer

CirrusLabs; Bengaluru; On-site; INR 4,000,000 - 7,000,000; Full time

Job summary

CirrusLabs is seeking a senior AI backend engineer with 7+ years of software experience to own backend features from design to production. You will work on AI/ML systems, agentic frameworks, and multi-step workflows, with emphasis on RAG architectures and vector search.

Proficiency in Python, FastAPI/Flask, and Kubernetes is essential. You will design scalable, observable systems, balance quality, latency, and cost, and collaborate with product and leadership while mentoring teammates.

Qualifications

• 7+ years of software engineering with strong backend ownership.
• Hands-on AI/ML engineering in production for LLM systems.
• Deep knowledge of agentic AI frameworks and multi-step workflow engines.
• Strong experience with RAG architectures and vector search tech (FAISS, Pinecone, Weaviate, pgvector, Milvus).
• LLM training and fine-tuning workflows (instruction tuning, domain adaptation, PEFT/LoRA/QLoRA).
• Python as primary language with FastAPI/Flask APIs and async patterns.
• Containerization & orchestration with Docker and Kubernetes (Helm, ingress, secrets, HPA).
• CI/CD ownership using GitHub Actions/GitLab CI/Jenkins; IaC (Terraform/Ansible/Helm).
• Datastores and caching (PostgreSQL, Redis) with exposure to NoSQL.
• Cloud fundamentals on AWS/GCP/Azure; IAM, networking, managed DBs.
• Observability stack (Prometheus, Grafana, OpenTelemetry).

Responsibilities

• Own complex AI backend features from design to production with measurable impact.
• Design reliable, observable, cost-aware scalable systems.
• Drive trade-offs between model quality, latency, and infrastructure costs.
• Mentor teammates and communicate with engineering, product, and leadership.

Skills

Backend ownership

AI/ML engineering

Agentic AI frameworks

RAG architecture

LLM training & tuning

Python

Docker & Kubernetes

CI/CD

Datastores & caching

Cloud fundamentals

Observability

Tools

FastAPI/Flask

Terraform/Ansible/Helm

GitHub Actions/GitLab CI/Jenkins

Prometheus/Grafana/OpenTelemetry

FAISS/Pinecone/Weaviate/pgvector/Milvus

TorchServe/Triton

Job description

• 7+ years of software engineering experience with strong backend ownership.


• Hands-on AI/ML engineering experience with LLM systems in production.


• Deep practical knowledge of Agentic AI frameworks and multi-step workflow engines.


• Strong experience with RAG architecture and vector search technologies (FAISS, Pinecone, Weaviate, pgvector, Milvus).


• Hands-on experience in LLM training and fine-tuning workflows (instruction tuning, domain adaptation, PEFT/LoRA/QLoRA, and parameter-efficient methods).


• Python (primary) with API frameworks such as FastAPI/Flask and asynchronous programming patterns.


• Containerization and orchestration with Docker and Kubernetes (Helm, ingress, secrets, HPA, resource quotas, monitoring).


• CI/CD ownership using GitHub Actions/GitLab CI/Jenkins and infrastructure-as-code patterns (Terraform/Ansible/Helm).


• Datastores and caching: PostgreSQL, Redis, and object storage, with exposure to NoSQL where needed.


• Cloud fundamentals on AWS/GCP/Azure (compute, container registries, IAM, networking, managed databases).


• Experience with observability stack (Prometheus, Grafana, OpenTelemetry, ELK/Opensearch, alerting).

Preferred / Plus

• Experience with MLflow, DVC, Weights & Biases, or equivalent experiment and dataset lifecycle tooling.


• Experience with serving stacks: Triton, vLLM, TorchServe, Text Generation Inference, BentoML, or equivalent.


• Experience with agent memory stores, retrieval quality benchmarking, and policy/safety layers for autonomous agents.


• Familiarity with Terraform, ArgoCD, or GitOps workflows for AI platform delivery.


• Security/compliance and governance awareness for AI systems (data privacy, prompt-injection controls, and auditability).

What We Expect

• Own complex AI backend features from design to production with measurable impact.


• Design systems for reliability, observability, and cost-aware scaling.


• Drive trade-off decisions across model quality, latency, and infrastructure costs.


• Communicate clearly with engineering, product, and leadership, and mentor other team members.