[Hiring] Senior AI Backend Engineer @BasharSoft

Role Description

Build scalable AI-powered backend systems that form the backbone of our client solutions. You'll architect robust APIs, integrate multiple AI providers with intelligent fallback strategies, and optimize real-time workflows that handle AI processing at scale.

What You'll Do

• Integrate and manage multiple AI providers (LLMs, image/video generation) with smart fallback and cost optimization
• Design and optimize prompt engineering strategies for consistent, high-quality AI outputs
• Build high-performance REST/GraphQL APIs using Django and FastAPI with async processing
• Develop real-time features using SSE, WebSockets, and background task processing (Celery)
• Architect scalable microservices with robust error handling, rate limiting, and monitoring
• Optimize database queries, implement caching strategies, and design efficient data models
• Implement RAG pipelines and vector embeddings for context-aware AI responses
• Write comprehensive tests, maintain CI/CD pipelines, and conduct code reviews

Qualifications

• 5+ Years of Experience
• Python & Frameworks: Django and/or FastAPI in production, async Python (asyncio, async/await)
• AI Integration: LLM APIs (OpenAI, Gemini, etc.), prompt engineering, token/cost optimization
• Databases: PostgreSQL (query optimization, indexing), Redis (caching, queuing)
• Async Processing: Celery or equivalent, message brokers (RabbitMQ, Redis, SQS)
• DevOps: Docker, Git, cloud platforms (AWS/GCP), CI/CD pipelines
• API Design: RESTful principles, authentication, rate limiting, error handling

Tech Stack

• Python 3.10+
• Django, FastAPI
• PostgreSQL, Redis, MongoDB
• OpenAI, Gemini
• Celery
• Docker
• AWS/GCP

Nice-to-Have

• Server-Sent Events (SSE) or WebSocket implementation
• Vector databases (Pinecone, Weaviate, pgvector) and RAG architectures
• MongoDB or NoSQL experience
• Multi-provider AI abstractions and failover strategies
• GraphQL, workflow orchestration (Temporal, Airflow), Kubernetes
• Observability tools (Datadog, Sentry, ELK)
• LangChain, LlamaIndex, or similar frameworks