The Fifth Elephant 2025 Winter Edition
2025-12-04..2025-12-06
Bengaluru, India
[The Fifth Elephant 2013, The Fifth Elephant 2015, The Fifth Elephant 2016, The Fifth Elephant 2018, The Fifth Elephant 2019, The Fifth Elephant 2020, The Fifth Elephant 2023 Monsoon, The Fifth Elephant 2025 Winter Edition, The Fifth Elephant Pune edition 2026]
| title | author | length |
|---|---|---|
| How Swiggy builds reliable agentic data co-pilots at scale | Amaresh Marripudi | 44m |
Production system walkthrough of how Swiggy turns plain-English questions into executable, auditable plans: agentic workflows that call validators and SQL… | ||
| the semantic layer your AI needs | Mayur Singal | 28m |
OpenMetadata claims to become the AI semantic layer by shipping an MCP server that ships with the install and exposes JSON‑schema metadata, lineage and auth so… | ||
| self-serve AI analyst | Prerit Khandelwal | 30m |
Razorpay’s in-house assistant, DataGaaru, combines a Reasoning Bot that generates and runs SQL against Trino with a Health Metrics Document Generator that runs… | ||
| The causality trap: why your LLM sounds smart but thinks shallow | Prasanna Bhogale | 40m |
LLMs pattern-match associations but cannot compute interventions: benchmark shows ~83% accuracy on associations but near-random on interventional/rung‑2… | ||
| Stop querying tables: give AI governed metrics instead | Nishant Bangarwa | 28m |
Claim: letting LLMs query raw tables produces hallucinated KPIs (example: AI reports average revenue per order $258 when dashboard shows much higher), misreads… | ||
| Synthetic data that meets enterprise compliance needs | Darshan Adiga, Lakshya Daulani | 23m |
An agentic LLM pipeline generates domain-rich, visually realistic enterprise documents using tool calls (charts/tables via image tools, open-source faker… | ||
| Mining digital exhaust | S. Anand | 02h 53m |
Workshop shows how to turn personal traces into concrete findings by actually scraping, coding and prompting: uses Chrome DevTools Protocol (CDP) + Playwright… | ||
| Decoding 1951 census: LLMs extract messy tables | Aryan Srivastava | 40m |
Digitizes India’s 1951 district handbooks across 40,000 pages by combining fine‑tuned OCR (Tesseract), layout detection (YOLO), and LLM orchestration, batching… | ||
| Live wireframing: lean BI using AI | Vikram Nayak | 02h 52m |
Claims a wireframe loop cuts dashboard rework and iteration cost by 100x cheaper and replaces guesswork with a six-step, repeatable process that uses LLMs to… | ||
| nothing matches | ||