Pranjul Rathour

Pitching at CSJMU, Kanpur Originally published at https://pranjulrathour.scult.in/blog/running-a-judges-only-debrief-after-the-event . That copy is the canonical version and gets updates first. A short debrief among the judging panel, right after results are finalised, is one of the most useful and most frequently skipped parts of running a hackathon. Everyone's tired and ready to leave, which is exactly why it gets cut — and exactly why the insight is still fresh enough to matter. What to actually cover in fifteen minutes Where the rubric felt genuinely ambiguous during real scoring, not just in theory. Any team or moment that tested the process in an unexpected way (a tie, a conflict of interest, a technical failure during a demo). One thing that worked well enough to repeat next time, not only what to fix. Why doing this immediately matters Waiting a week to gather this feedback loses the specific, concrete detail that makes it useful — by then it's vague impressio...

At an Integral Startup Foundation hackathon Originally published at https://pranjulrathour.scult.in/blog/evaluating-fine-tuning-roi-was-it-worth-it . That copy is the canonical version and gets updates first. A fine-tuning project that ships doesn't automatically mean fine-tuning was the right call. It's worth honestly comparing the result against the simpler alternatives that were available from the start. The comparison to actually run Same eval set, three ways: the base model with a good prompt, the base model with few-shot examples, and the fine-tuned model. Measure quality difference alongside cost difference — a marginal quality gain that cost forty GPU-hours may not have been worth it against a well-engineered prompt. Factor in maintenance: a fine-tuned model needs retraining as requirements evolve; a prompt can be edited in seconds. When the ROI question has an easy answer If the task structurally can't be solved by prompting at all — a fixed classificatio...

Carousel · 8 slides Three first prizes, one year, one loss that taught more. Feb — Changethon, IIT Roorkee: Krishi Gyan, an AI agribot (React, Flask, Gemini Flash). Apr — Product Genesis, CSJMU: BrandHive, a pitch. Jun — BYTEBATTLE by MeetKats: Annapurna, surplus-food coordination (React, Supabase), vs 15+ teams. Dec — STROT, 48 hours, didn't win. Lesson: clarity and storytelling matter as much as the system. Same team every time: Raman Shukla, Saksham Gupta and me. Open to judge hackathons and give this as a talk, free for student communities. Pranjul Rathour · GenAI engineer, Kanpur · pranjulrathour41@gmail.com · pranjulrathour.scult.in/invite Slide 1 of 8 Slide 2 of 8 Slide 3 of 8 Slide 4 of 8 Slide 5 of 8 Slide 6 of 8 Slide 7 of 8 Slide 8 of 8 Pranjul Rathour GenAI engineer, Kanpur · 3x first-prize hackathon winner · campus mentor I ship production RAG pipelines, fine-tune LLMs and build agentic AI products end to end. I lead engineering at SCULT INDIA for a 14-mem...

Carousel · 9 slides Broadcom may lend Anthropic up to $42 billion, per Anthropic's IPO prospectus (reported 1 Oct 2026). → Convertible notes that could turn into Anthropic shares → Covers about one-third of the $125.2B Anthropic pledged for 5 years of TPU compute → Broadcom can bring in a financing partner; no notes expected before the IPO → The prospectus flags "potential conflicts of interest": Broadcom is both supplier and lender Chipmakers are now financing their own buyers. My take for builders: stay model-agnostic. I'm Pranjul Rathour, GenAI engineer from Kanpur. Want me to judge your hackathon or speak at your college? pranjulrathour.scult.in/invite Slide 1 of 9 Slide 2 of 9 Slide 3 of 9 Slide 4 of 9 Slide 5 of 9 Slide 6 of 9 Slide 7 of 9 Slide 8 of 9 Slide 9 of 9 Pranjul Rathour GenAI engineer, Kanpur · 3x first-prize hackathon winner · campus mentor I ship production RAG pipelines, fine-tune LLMs and build agentic AI products end to end. I lead...