← sources

[Feedback Requested] Planning a "Research-First" ML Cohort for Undergrads. Is this actually needed?

source: reddit (reddit:developersIndia)published 2026-09-27 07:04ingested 2026-09-28 02:38

retained text

Hi everyone,

I am seeking honest feedback on a community/course initiative I plan to launch for Indian undergraduate first-year students.

**The Context:** I believe the current education landscape is saturated with "zero to hero" coding boot camps and learn AI in 7 days tutorials. While these are great for getting started, I often find that students lack the deep, theoretical foundations required for actual research or heavy engineering roles later in their careers.

I want to build a small community (cohort-style) to bridge this gap, but before I invest the time, I want to know if I'm solving a real problem or just adding to the noise.

**My Background**

* **Current:** Fully funded Graduate Researcher in Germany.
* **Past:** 2+ years as an ML Scientist (Applied AI Research org) and 1 year as a Research Associate.
* **Academic:** 3+ Top-tier publications.

**The Curriculum Idea**: Instead of teaching library imports (sklearn/torch), I want to focus on he "boring" but essential foundations:

1. **Mathematics for ML:** Heavy focus on Linear Algebra & Calculus (Manual derivations).
2. **Probabilistic & Statistical ML:** Understanding uncertainty, distributions, and estimation.
3. **ML Theory:** Generalization, Bias/Variance trade-offs, VC Dimension (Intro).
4. **Deep Learning:** Building neural networks from first principles.
5. **Research Capstone:** Literature review + Benchmarking + A deep research project.

**The Filter Mechanism**: I want this course to be free, but I want to avoid tourists who join and drop out in Week 2.

* **The Model:** A token fee of **1000 INR**. (or less) \[FREE!\]
* **Refund Policy:** A 100% refund is available if the student completes all assignments.
* **Financial Aid:** The fee is waived entirely for students with genuine financial constraints (based on trust).
* **The Constraint:** Assignments must be completed without the use of AI tools (such as ChatGPT/Copilot). If a student uses AI to bypass the learning process, they forfeit the deposit (donated to charity) and are dropped.

**My Questions for the Community**

1. **Do you know if this is actually needed?** Are there already enough high-quality, free, community-driven resources for *theoretical* ML?
2. **Is the curriculum too aggressive?** Is this too much for Freshmen (1st/2nd years) to handle alongside college?
3. **The Deposit:** Is the refundable model a good psychological trigger for commitment, or does it look suspicious/scammy coming from an individual?

Thanks in advance for your thoughts.

\---  
Note: The post is AI-Gen for clear communication and brevity.  


\--  
Edit: It will be free.   
I plan to do this mentorship program (and maybe would ask my friends to join in along the way - to share their experiences in ML so far). I created this course page, and a simple placement check for you to figure out if this course is going to be a good fit for you The test ended up being a bit too difficult, but try to do whatever you can. Thanks Link: [https://blankpageml.tech/](https://blankpageml.tech/)

document fetched 2026-09-28 00:25 · canonical https://reddit.com/r/developersIndia/comments/1wrdg4y/feedback_requested_planning_a_researchfirst_ml