AutomatR
Website:
automatr.tech
Job details:
Role: Lead AI Engineer - AI Assistants & Agentic AI
Exp: 2+ to 5 Years
We are looking for a highly experienced and hands-on Lead AI Engineer who has built and deployed real-world AI Assistants & Agentic AI and understands deeply how they work internally — including memory management, tool usage, reasoning loops, context orchestration, and multi-agent coordination.
This role requires someone who does not just experiment with LLM APIs but has architected production-grade AI assistants capable of:
· Tool calling / function execution
· Context management & long-term memory
· Retrieval-augmented reasoning
· Goal-based task planning
· Autonomous decision-making
· Multi-step workflow execution
You will lead the design and evolution of next-generation AI Assistants integrated into enterprise automation systems.
What You Will Do
1. Build Advanced AI Assistants
· Design and implement production-grade AI Assistants.
· Develop:
o Tool-augmented agents
o Multi-step planners
o Self-reflective reasoning systems
o Memory-enabled assistants (short-term + long-term)
· Implement function calling, tool orchestration, and action chaining.
· Build assistants capable of interacting with APIs, databases, and enterprise systems.
2. Deep Understanding of Assistant Internals
· Design context window management strategies.
· Implement:
o Conversation memory layers
o Persistent vector-based memory
o Context compression strategies
· Reduce hallucinations via:
o Grounded retrieval
o Tool validation
o Guardrails
· Architect reliable assistant behavior in enterprise settings.
3. Agentic & Multi-Agent Systems
· Design goal-driven AI agents.
· Build multi-agent workflows using:
o LangGraph
o LangChain
o LlamaIndex
o CrewAI (or similar)
· Implement:
o Task decomposition
o Agent-to-agent communication
o Delegation & planning loops
· Improve agent determinism and traceability.
4. RAG & Knowledge Systems
· Architect scalable Retrieval-Augmented Generation pipelines.
· Work with vector databases (ChromaDB, FAISS, Weaviate, Milvus).
· Implement hybrid search and reranking.
· Design structured & unstructured ingestion pipelines.
5. LLM Optimization & Fine-Tuning
· Fine-tune and optimize Small/Tiny LLMs (Phi-3, Mistral, Llama 3, etc.).
· Apply LoRA, QLoRA, PEFT techniques.
· Optimize inference for low-latency AI assistants.
· Implement model routing and fallback strategies.
Required Skills:
· 3–6 years in AI/ML/NLP.
· Strong expertise in LLMs and transformer-based architectures.
· Hands-on experience building AI Assistants in production.
· Deep understanding of:
o Tool-calling agents
o Memory management
o RAG pipelines
o Context engineering
· Proficiency in Python.
· Experience with:
o Hugging Face
o PyTorch
o LangChain / LangGraph / LlamaIndex
o Vector databases
· Experience deploying AI systems using Docker/Kubernetes.
What We’re Specifically Looking For
We want someone who can confidently answer:
· How does an AI Assistant manage context internally?
· How do you prevent hallucinations in tool-calling agents?
· How do you design long-term memory for assistants?
· How do multi-agent systems coordinate tasks?
· How do you make assistants reliable in production?
This is not a prompt-engineering role. This is a systems-level AI engineering role
Click on Apply to know more.