Overview
We are seeking an
Agentic AI Developer
with hands-on expertise in building multi-agent systems using
LangChain, LangGraph, and LangSmith
. The successful candidate will design and deploy a
3-agent research team
with a
human-in-the-loop architecture
, deployed on
AWS AgentCore
.
This project requires strong experience with
Python
,
Pydantic v2
, and validation-driven AI workflows. The agents should provide outputs in
OpenAI Completion API style
, maintain
persistent memory
, and enforce
schema validation
against defined
Pydantic models
.
Key Responsibilities
* Architect and implement a
3-agent research system
(collaborative/parallel agents) with human-in-the-loop oversight.
* Develop agent workflows leveraging
LangChain, LangGraph, and LangSmith
for orchestration, monitoring, and debugging.
* Ensure agents produce outputs consistent with the
OpenAI Completion API format
.
* Implement
persistent memory layers
for agent context retention across sessions.
* Validate all agent outputs against
Pydantic v2 schemas
to guarantee structured and reliable responses.
* Deploy and optimize the system on
AWS AgentCore
, ensuring scalability, reliability, and secure operations.
* Collaborate with stakeholders to define success metrics and finalize system deliverables.
Required Qualifications
* Proven experience
with
LangChain, LangGraph, and LangSmith
in production or advanced prototyping settings.
* Strong
Python
development background, including advanced async patterns.
* Expertise in
Pydantic v2
for schema definition, data validation, and error handling.
* Practical experience deploying AI/agent systems into
AWS environments
(preferably
AWS AgentCore
).
* Knowledge of
LLM orchestration patterns
(multi-agent workflows, memory persistence, validation pipelines).
* Familiarity with
human-in-the-loop
design patterns for AI systems.
Nice-to-Have Skills
* Experience with
OpenAI APIs
(Completion & ChatCompletion endpoints).
* Set up streaming responses.
* Familiarity with
observability tools
(e.g., tracing, logging, and monitoring agent interactions).
* Prior work with
agent-based research systems
or
RAG pipelines
.
* Understanding of
security best practices
for AI deployment in cloud environments.