Inspiration
The process of internal recruitment has seen little innovation the traditional approach has many problems like slow processing, unconscious bias, skill gap identification etc. With the advent of Gen AI & AI Agents we can slow these issues and optimize the process to proactively complete internal recruitment.
What it does
CareerCraft AI is an agentic solution it employs specialized agents to perform specific tasks. It has a specialized planner agent that handles the orchestration. There are specialized agents like Profile search to search candidates best suited based on the JD. Profile matcher to assess and rank candidates based on past performance and course recommender for provideing courses that aim to fill gaps in candidates skill.
How we built it
We built CareerCraft AI as a planner‑led, tool‑rich agent —running on cloud agent tooling (AWS Bedrock Agentcore)—that parses JDs, searches profiles, ranks matches, and recommends upskilling via HR/LMS/PMS connectors, all instrumented for traceable orchestration and measurable hiring impact.
Challenges we ran into
The main challenges we ran were integration of agents with MCP server & orchestrating the flow between with a2a protocol.
Accomplishments that we're proud of
End to end deployment on AWS.
What we learned
Understanding and implementation of new age technologies like Agentic AI, MCP and a2a .
What's next for Careercraft AI
- Integration with collaboration tools like slack, zoom.
- Changes to the UI to provide list of all candidates in a traditional UI.
Built With
- a2a
- amazon-web-services
- apprunner
- bedrock
- fastapi
- mcp
- python
- react
- strands
Log in or sign up for Devpost to join the conversation.