AI Architect - Multi-Agent Systems
First seen by Alion on Sep 10, 2026.
Experience Level : 8+ years (minimum 2-3 years in ML, Gen AI and Multi-Agent Systems)
Key Responsibilities :
- Architect and implement agentic AI systems using modern LLM orchestration frameworks (LangChain, CrewAI, AutoGen, etc.).
- Design multi-agent collaboration models including planner-solver, autonomous teams, and goal decomposition agents.
- Build reusable tooling, APIs, and memory architectures for agent interaction, coordination, and context persistence.
- Lead hands-on development and deployment of GenAI applications (e.g., assistants, copilots, decision support).
- Evaluate and integrate LLMs (OpenAI, Claude, Mistral, LLaMA, etc.), vector databases (Pinecone, Weaviate, FAISS), and retrieval systems (RAG).
- Optimize agent performance for real-time environments, reliability, scalability, and ethical constraints.
- Guide teams in adopting agent frameworks, best practices, prompt engineering, and model fine-tuning.
- Collaborate with stakeholders to translate business requirements into technical solutions using agent-based paradigms.
- Continuously monitor trends in multi-agent systems, cognitive architectures, and open-source AI frameworks.
Must-Have Skills :
- 2+ years of hands-on experience in agentic AI / multi-agent systems.
- Proficiency with LangChain, Langraph, CrewAI, AutoGen, Haystack, or equivalent frameworks.
- Strong background in Python and experience with prompt engineering, tools integration, and chaining logic.
- Solid understanding of LLM APIs, RAG, vector stores, tool use, and memory architectures.
- Hands-on experience with open-source and commercial LLMs (e.g., GPT-4, Claude, Gemini, Mistral).
- Experience deploying AI agents in cloud-native environments (AWS, GCP, Azure).
- Ability to lead architectural discussions, PoCs, and hands-on development in fast-paced environments.
- Model-cost profiling and budgeting (API call minimization, batch vs. streaming).
- Latency tuning for real-time agents, Autoscaling strategies.
Good-to-Have Skills :
- Exposure to Autonomous AI agents (AutoGPT, BabyAGI, CAMEL, MetaGPT).
- Understanding of LLM fine-tuning, adapters, and RLHF.
- Experience with agent simulation, environment modelling, or reinforcement learning is a plus.
- Familiarity with compliance, privacy, and safety in GenAI deployments.
- Prior experience in building domain-specific agents (Lifescience, healthcare, Pharma).
Skills
Machine Learning, Artificial Intelligence, LangChain, Generative AI, Agentic AI, LLM, GPT