Backend Engineer (AI & Infrastructure)

Backend Engineer (AI & Infrastructure) Full-time .

Engineering .

1–3 Years Experience .

Onsite/ Hybrid

About the Role We are looking for a curious and hands-on Backend Engineer who sits at the intersection of robust systems engineering and applied AI. You will design and build scalable backend services architect AI-powered agent workflows and own the infrastructure that keeps them running reliably in production. This is an early-career role with high ownership ideal for someone who loves going deep on low-level design as much as they love experimenting with LLMs and vector databases.

Backend Engineering Design and build scalable REST and gRPC APIs using Python or Go Apply solid low-level design patterns: SOLID principles design patterns clean

architecture Work with relational (PostgreSQL MySQL) and NoSQL (MongoDB DynamoDB)

databases Write efficient testable and well-documented code with async/concurrent

programming Participate in code reviews and contribute to engineering best practices

Distributed Systems & Real-Time Communication Build event-driven architectures using Apache Kafka for high-throughput data

pipelines Implement caching and pub/sub messaging with Redis Develop real-time features using WebSockets and Server-Sent Events (SSE) Work with message queues such as RabbitMQ or AWS SQS Apply rate limiting throttling and backpressure strategies in production systems

AI Agents & Large Language Models Integrate with LLM APIs (OpenAI Anthropic Claude Google Gemini) for

production use cases Build autonomous agent workflows using patterns like ReAct Plan-and-Execute

and multi-agent hierarchies Design and manage tool/function calling context window budgeting and memory

strategies

Develop Retrieval-Augmented Generation (RAG) pipelines end-to-end Work with agent frameworks such as LangChain or LlamaIndex Apply prompt engineering techniques for reliability and consistency

Vector Databases & Semantic Retrieval Hands-on experience with vector databases: Pinecone Weaviate Qdrant or

ChromaDB Understand embedding models and how to select generate and index them

effectively Implement chunking indexing and hybrid search strategies (BM25 dense

retrieval) Design retrieval pipelines optimised for accuracy latency and cost

Infrastructure & Cloud Containerise services with Docker and understand Kubernetes for orchestration

basics Hands-on experience with at least one major cloud provider: AWS GCP or

Azure Work with serverless compute (Lambda Cloud Functions) and managed services Exposure to Infrastructure as Code tools such as Terraform or AWS CDK

CI/CD & Developer Experience Build and maintain CI/CD pipelines using GitHub Actions or GitLab CI Manage Docker image builds container registries and automated deployments Apply deployment strategies: blue-green deployments canary releases and

rollback procedures Handle secrets and environment configuration management securely Write automated tests (unit integration end-to-end) as part of the pipeline

Nice to Have Experience with observability tooling (Datadog Grafana OpenTelemetry)

especially for LLM latency and token cost monitoring Familiarity with guardrails output validation and safety layers for AI agents Exposure to fine-tuning or LoRA workflows on open-source models Knowledge of streaming architectures: chunked responses token streaming to

the frontend API security best practices: OAuth 2.0 JWT RBAC Open-source contributions or personal AI projects on GitHub

What We Value Ownership mindset you take a problem from an ambiguous brief to production

without hand-holding Clear written communication especially for async and distributed teams Opinionated about code quality and design but pragmatic about trade-offs Stays current with the fast-moving AI ecosystem without getting distracted by

hype Curiosity and a habit of building things to learn side projects experiments and

open source count

Experience range: 1–3 years. We care more about depth and curiosity than years on a résumé. A strong GitHub profile or portfolio with real-world projects will weigh heavily in our evaluation.