Loading page content…
AI-Focused Society
Beyond theObvious.
A college-wide community for everyone interested in AI, software, data, systems and the ideas shaping tomorrow.
The society
About CORTEX
A community for students across the college to explore AI, software, data, systems and emerging technology together.
- Technical articles and discussions
- Curated research reading
- Structured learning roadmaps
- Problem-solving challenges and team projects
1Community members
3Divisions
Mission
Promote peer learning, practical AI engineering, career guidance, industry connections and accountable, faculty-guided project work.
Vision
A college-wide community where students from every branch and year learn together, build useful projects and grow into thoughtful technologists.
Philosophy
Learn by doing, share knowledge, include students at all skill levels, and work under faculty guidance.
Structure
Our divisions
The proposal lists three divisions: Technical, Events and Outreach & Marketing.
01
Technical
Two Chapter Heads and associates
Proposed: designs learning tracks and technical sessions, mentors project groups, reviews code and maintains shared notes and resources.
View members02
Events
Two Chapter Heads and associates
Proposed: organises hackathons, workshops and competitions; coordinates speakers, venues, schedules, registrations and permissions.
View members03
Outreach & Marketing
Two Chapter Heads and associates
Proposed: manages social media and community communications, creates promotional materials, onboards members and promotes club activities.
View membersCORTEX calendar
Upcoming events
Talks, workshops and society sessions. Register directly on the CORTEX website.
New event dates will appear here once published. View the event calendar
People
Member directory Leadership & team
Meet the people helping shape CORTEX and explore the interests they choose to share.
Featured: Manav TuliPresident
CORTEX Insights
All articles Writing from members
Technical articles, with questions and answers in the open.
A practical reading path for coding agents
A source-linked route from Git internals to agent architecture, with notes on measurement, safety and the limits of company-specific case studies.
Research Radar
Browse Radar Worth your reading
Curated papers and engineering write-ups, linked to their original sources with our own summaries.
Components of a Coding Agent
Sebastian Raschka
A technical guide to the parts that make up a coding agent, including its model, context, tools and control loop.
Why read it: Build a mental model of agent architecture before focusing on any one coding product.
Redesigning source control at Anthropic
Anthropic · 13 Oct 2026
Anthropic describes building Scion as its infrastructure needs changed. It reports about 2,500 code changes merged per day in May 2026 and more than 30,000 per day in September; this is a company-specific case study, not a general baseline.
Why read it: See how a team reasons about source control bottlenecks when its workload changes sharply.
When AI builds itself
Anthropic Institute · 4 Jun 2026
Anthropic reports that its typical engineer merged 8× as much code per day in Q2 2026 as in 2024. The article cautions that lines of code measure quantity, not quality, so the figure is not a direct measure of productivity.
Why read it: A first-party account of how AI changes engineering work, with an explicit discussion of what its productivity measure can and cannot show.
Prerequisites: Basic familiarity with software engineering workflows.
Learning Hub
All roadmaps Structured roadmaps
Ordered stages, curated materials and personal progress tracking. Free to read; sign in to track your own progress.
From Git to Coding Agents
A self-paced, source-linked path from Git internals to AI coding agents, ending with an optional read-only repository assistant challenge. Reading time is an estimate; one linked article is subscriber content.
Problem Lab
All challenges Challenges worth solving
Member-proposed problems, reviewed by the society. Express interest, form a team and ship something real.
Build a read-only repository assistant
Create a small assistant that answers questions about a public or synthetic Git repository and cites the exact file paths and line ranges behind each answer. A team can divide the work across research, interface design,…
repository parsingretrievalcitations
Join CORTEX
Become a member
CORTEX is a college-wide community for students interested in AI, software, data and technology. Students from every branch are welcome to learn; Community membership is a separate selective pathway with auditions. Prior technical experience is not required.
Applications open after a session ends as third-year members move into fourth year, or when a Community position becomes vacant. Selection includes auditions.Create account
Membership benefits
- Write and publish approved technical articles
- Propose challenges for the Problem Lab
- Form project teams across different interests and divisions
- Collaborate with other CORTEX Community members
How it works
- Create an account and verify your email
- Complete your profile
- Apply when the administrator opens a selection window after a session ends as current third-year members move into fourth year, or when a Community position is vacated
- Take part in the audition round
- CORTEX will contact you with the selection outcome