We are looking for a Backend AI Engineer who builds production-grade systems powered by Large Language Models. LLMs are not magic they are probabilistic components that must be engineered carefully inside reliable backend systems. This is a systems engineering role, not just prompt engineering.

You are a backend engineer first and an AI systems builder second. You think in terms of systems, reliability, and architecture, understanding that LLMs are just one component in a larger backend system. You have built real-world APIs, agents, and automation workflows, and you care deeply about latency, reliability, debugging, and scalability.

Key Responsibilities:

1. LLM Agent Engineering:

• Design stateful LLM agents using LangChain /LangGraph
• Build robust tool-calling systems with strict JSON schemas
• Handle hallucinated tool inputs, incorrect reasoning,retries
• Build RAG pipelines using vector databases
• Experiment with prompt strategies and multi-step reasoning

2. Backend System Development:

• Build scalable APIs using FastAPI
• Design backend architecture: session management, state handling
• Work with PostgreSQL, Redis, and Vector databases
• Write clean, typed, production-ready Python services

3. Real-Time & Voice Systems:

• Integrate with telephony/voice infrastructure (Twilio, etc.)
• Optimize ASR LLM TTS pipeline for low latency
• Handle barge-in, silence detection, background noise
• Implement turn-taking logic for real-world voice edge cases

4. Evaluation & Observability:

• Build eval systems using LangSmith / DeepEval / custom
• Measure tool accuracy, task completion, hallucination rate
• Implement logging, tracing, and debugging for agent workflows
• Optimize latency, token usage, and LLM cost efficiency
  

5. External Tooling & Integrations:

• Integrate Google Calendar, HubSpot, Salesforce
• Handle OAuth authentication and API rate limits
• Build reusable tool registry architecture for agents
• Implement retries and failure recovery strategies

Technical Requirements:

1. Backend Engineering:

• 3+ years of Python development experience
• Strong hands-on experience with FastAPI
• Skilled in async systems and building scalable REST APIs
• Solid understanding of backend system design and architecture

2.  AI / LLM Systems:

• Practical experience with LangChain and LangGraph
• Proficiency in LangSmith for tracing and evaluation
• Experience building RAG (Retrieval-Augmented Generation) systems
• Knowledge of tool/function calling with LLMs
• Experience in prompt iteration, testing, and evaluation workflows

3. Data Systems:

• Strong working knowledge of PostgreSQL (relational databases)
• Experience with Redis for caching, queues, and state management
• Familiarity with vector databases such as Pinecone, Weaviate, or similar systems

Nice to Have:

• Exposure to multi-agent systems
• Prompt versioning & experimentation
• LLM cost optimization experience
• Streaming systems / WebSockets
• Production observability in AI systems
  

Other Details:

• Job Timings: 9 working hours between 12 pm and 12 am.
• Office location: Off to Shahrah-e-Faisal, PECHS, Karachi