Live industrial project · 2026 · Applications open

Solumn AI Foundations Programme

Evaluating the safety of frontier AI systems.
A six-week applied AI research engagement. Places are filled as candidates qualify.

6 weeksFrom your date of joining
₹2,00,000Stipend for the engagement
RemoteHours around your academic timetable
PPOOffered on merit at close

The work

Work on live AI-safety problems.

You will join one live workstream and build something that is reviewed, graded and used — not a classroom exercise.

01

Agentic environments

Containerised tasks and deterministic graders for evaluating agent behaviour under pressure.

02

Vulnerability datasets

Security tasks from real CVEs, backed by executable verification.

03

Adversarial evaluation

Tool-use and MCP safety evaluations against realistic attack conditions.

Who can apply

Final-year technical students.

B.Tech · Dual Degree · M.Tech

CSE · ECE · EE · Mathematics & Computing

Evidence of building matters.

Selection process

CV submissions close
Assignment shared with shortlisted applicants
Assignment submissions close
Interviews

Apply to Foundations 2026.

Keep it concise. We care most about evidence of things you have actually built.

1. Your details

2. Academic background

3. Evidence of building

Link to your strongest work: repositories, open source, security, systems, benchmarks or research.

4. Availability

Questions? research@solumn.ai