I’m Rashmi Shaw, open to early-career Backend, Applied AI, and Agent Systems roles worldwide. Remote or willing to relocate.
I co-authored a merged NVIDIA Model Optimizer fix and build agent infrastructure with Python, FastAPI, and LangGraph.
https://t.co/HVlRetPSDn
Built AI agents.
Worked on backend systems.
Contributed to open source.
Still struggling to land my first full-time role.
What actually matters more for freshers in 2026: DSA, projects, referrals, or luck?
People say, "Contribute to open source. It helps you get hired."
I did.
Merged PRs. Real code. Real repositories.
But getting interviews is still difficult.
Has open source actually helped anyone here land a job?
One thing I wish more startups did:
Give freshers a real engineering problem to solve instead of filtering only by experience.
I'd happily demonstrate my backend/AI skills through actual work.
Any founders hiring builders?
700+ cold emails. Hundreds of applications. Projects built. Open-source contributions.
Still no job.
Graduated in 2026, and honestly, I don't know what I'm missing anymore.
Not looking for sympathy. Just an opportunity to prove myself.
https://t.co/HVlRetPSDn
@HashgraphOnline Good question! MCPShield checks the JSON-RPC envelope, then runs policy on tools/call before forwarding to the tool server for schema validation. Denials include the policy name and reason. Calls pending approval return an approval ID. The matched rule ID stays internal.
Who gave the AI agent permission to do that?
I built MCPShield so teams can check an MCP tool call against policy before it runs and review what happened afterward.
https://t.co/XfwhTELbTY
“Where did that answer come from?” should take one click.
I built NexusAgent to search code, docs, and tickets and return answers with citations you can follow.
I’m open to early-career Backend and Applied AI roles in India. DMs open.
https://t.co/aejC4xMuYN
The MCP tool was registered. The client still couldn’t find it.
I helped investigate why OpenEnv’s mode-specific tools were missing over JSON-RPC and contributed regression tests for the upstream fix.
https://t.co/XDQb7CHvzR
The human-in-the-loop demo approved its own draft.
My PR merged into Temporal’s Python samples. Now you can approve a draft or ask for revisions from your terminal. Both LangGraph demos run without an LLM API key.
https://t.co/A94bT1L7WK