IIT-BHU students preparing for Confluentia have several different participation routes to check this week, rather than a single competition deadline. The second edition of the industry-academia gathering is scheduled for October 24 and 25, 2026, at Swatantrata Bhawan, with hackathon activity beginning before the main programme.

The Student Alumni Interaction Cell and Alumni Association of IIT-BHU are organising the event. Its published programme includes research and innovation displays, industry-led workshops, alumni mentoring and student project opportunities. The organising team’s October 11 announcement also promoted activities across October 23 to 25; individual programme calendars provide the more precise deadlines.

Applied ML teams face an October 21 submission

The Tredence Applied ML track lists its second mentor review for October 20 and final submission for October 21. Problem statements were released on October 17. Tredence judges are scheduled to evaluate submissions on October 22 and 23, before 10 to 15 shortlisted teams present offline on October 24.

The challenge covers generative AI, AI agents, computer vision, multimodal AI, optical character recognition and speech. Teams select a problem and receive an assigned Tredence mentor. The programme lists two progress reviews, on October 18 and 20, with the first described as a five-to-ten-minute check-in in slots set by the organisers.

Participation is restricted to IIT-BHU students, with one to three members per team and a ceiling of 100 teams. The page says registration closes when that limit is reached. It advertises cash prizes, goodies and internship offers, but does not specify prize amounts. A registration link alone does not confirm that places remain.

A separate 48-hour reinforcement-learning challenge

The reinforcement-learning track, run with the PhysicsWallah initiative 2une, follows a different timetable. Its problem statements are due on the morning of October 23, starting a 48-hour build period. Submissions close on October 25, and PhysicsWallah mentors are scheduled to hold a mentoring and doubt-clearing session on October 24.

This track is also limited to IIT-BHU students in teams of up to three. The programme describes reinforcement learning as training agents through rewards rather than labels. That is the stated focus of this competition; it should not be confused with the separate Applied ML track’s wider set of problem areas and earlier submission date.

The portal says students can use one Confluentia account for both hackathon tracks and other activities, but the two tracks require separate registrations. For the reinforcement-learning challenge, registered teams receive the problems through their dashboard and can confirm attendance at the mentoring session there.

Project work does not guarantee a paid internship

The Internship Preparation League offers another route: companies propose real problems, students submit plans, and selected participants undertake projects normally lasting four to eight weeks. Company-student meetings are scheduled for October 24. The public programme page, checked on October 20, still says student applications will open soon, so the published guide should not be read as proof that applications are already live.

The guide allows remote or hybrid work, with duration and scope agreed between the company and students. A stipend during the project is optional. Paid internships may follow for strong performers, depending on the company’s needs; their number, pay, duration and working mode are not guaranteed. Completion involves agreed deliverables and a final review.

Students may propose solutions to several problems, but shortlisting has a consequence: their proposals disappear from other companies’ lists. If released, those proposals return. The guide says companies initially see academic and professional profile information; email addresses and phone numbers become available after shortlisting.

Alumni mentoring uses its own forms

Match Circles pairs students with IIT-BHU alumni whose experience fits their interests. The organisers’ October 11 announcement named areas including AI and machine learning, software development, quantitative work, data science and product management. The format is a face-to-face conversation over a meal, rather than a project selection or competition.

The current Match Circles page provides separate Google Forms for students and alumni. Students describe their interests, goals and subjects they want to explore, after which the organisers arrange a match. Its student route is restricted to IIT-BHU. The page describes a networking lunch, but does not publish a specific lunch time or a remaining-place count.

These are distinct from the institute’s earlier Traction entrepreneurship programme and its forthcoming Gleeble materials-testing workshop, both covered separately by Dalimss News. Confluentia participants should use the relevant programme page rather than assume that registering for the summit also enters every competition or mentoring activity.