Submissions
The Fifth Elephant
The Fifth Elephant

The Fifth Elephant 2025 Winter Edition Call for Proposals

On AI disrupting analytics - make a submission

Accepting submissions

Not accepting submissions

Speaking policies and guidelines One speaker per talk. No co-speakers or joint presenters. Vendor-neutral submissions. Talks must focus on technical concepts, case studies, and real-world learnings. If your intention is to showcase a product, platform, or tool, you’re welcome to submit a (paid) pit… expand

Speaking policies and guidelines

  • One speaker per talk. No co-speakers or joint presenters.
  • Vendor-neutral submissions. Talks must focus on technical concepts, case studies, and real-world learnings.
  • If your intention is to showcase a product, platform, or tool, you’re welcome to submit a (paid) pitch session, or run a sponsored workshop to teach your tool/product/platform. (Reach out to sales@hasgeek.com for sponsored workshops and pitch sessions.)
  • We strongly encourage talks that involve open source technology, tools, or frameworks - especially those built or adopted within your organization.
    The goal is to ensure the audience can take back learnings and technology they can experiment with or implement independently.

Guidelines for submissions

BEFORE you begin writing your submission, please give some thought to the following:

  • Who is the audience for your session? Think about their interests, work roles, challenges, age or experience as you decide this.
  • What problem/pain are you trying to solve (for the audience)? This should be something that is communicated clearly so that they have a sense of your session’s importance.
  • What will be the scope of your session? This will help identify the central topic or theme and should describe broad areas you plan to cover during the session?
  • How will participants benefit from your session? Think of practical and specific ways in which they will be able to apply the knowledge they gain, and beyond just general awareness.
  • What is the appropriate format for your session, given the audience and objectives that you have in mind?

The most successful talks and sessions are those where presenters are able to abstract an actionable insight from a common pain area, enlighten the audience about something new, provide a fresh perspective, and/or demonstrate innovation.

Here’s a guide for speakers to draft their presentations.

Make a submission
Sandhiya Giri

Sandhiya Giri

From Queries to Conversations: Building an Agentic Analyst for Investment Performance Insights

Session description In the world of investment performance analytics, teams are overwhelmed by repetitive ad-hoc data requests each requiring manual SQL, custom calculations, and cross-checking across multiple systems. This session introduces an Agentic Performance Analytics system: a virtual analyst that turns natural language questions like “What’s Fund X’s 3-year return?” into accurate, audita… more
  • 3 comments
  • Submitted
  • 20 Oct 2025
Type of submission: 30 mins talk

Nishant Bangarwa

Stop Prompting Your Database: Give AI a Metrics Layer Instead

As AI becomes central to analytics workflows, many teams try to let LLMs query databases directly—only to find that raw tables lack the semantics AI needs. The result: hallucinated KPIs, ambiguous joins, inconsistent definitions, invalid SQL, and costly full-table scans. more
  • 3 comments
  • Submitted
  • 27 Oct 2025
Type of submission: 30 mins talk

Mayur Singal

OpenMetadata, Bringing Semantic Understanding to Your Data to Power AI

AI can generate code, queries, and dashboards but it still doesn’t understand your data. Without context, lineage, ownership, and business meaning, AI systems remain disconnected from the reality of your enterprise data. more
  • 1 comment
  • Submitted
  • 03 Nov 2025
Type of submission: 30 mins talk

Marripudi Amaresh

Building Data Co‑Pilots that work!

Building Data Co‑Pilots is a session on how to use agentic LLMs to reliably answer complex business questions by aggregating data across many sources—warehouse/lakehouse tables, metrics/semantic layers, feature stores, logs, and knowledge bases. We will walk through the system design that turns a plain‑English question (e.g., “Why did cancellations spike in City X yesterday?”) into an executable … more
  • 1 comment
  • Submitted
  • 04 Nov 2025
Type of submission: 30 mins talk

prerit Khandelwal

From Bottleneck to Self-Serve: How Razorpay Built and Scaled its AI Analyst, DataGaaru

At Razorpay, the growing demand for data & insights was increasing manual dependence, slowing down critical business decisions. This led us to ask a question: “Are chatbots the future of analytics?” Our answer is DataGaaru, a home-grown AI Analyst Assistant that is fundamentally reshaping how we interact with data. Today, it handles over 1,500 queries monthly and has already reclaimed 20% of our … more
  • 1 comment
  • Submitted
  • 08 Nov 2025
Type of submission: 30 mins talk

Aryan Srivastava

Extracting Data from Historical Documents: Harnessing LLMs to Parse Format-Variant Tables at Scale

Traditional OCR systems are excellent at recognizing text but fall short in understanding the structure of complex historical documents—particularly tables with varying column arrangements, inconsistent layouts, and annotations. This structural ambiguity often makes it infeasible to extract reliable, machine-readable data from large collections of such documents. more
  • 2 comments
  • Submitted
  • 09 Nov 2025
Type of submission: 30 mins talk

Rajesh RS

Building Self-Discovering Analytics with Praval, an open-source Agentic AI framework

Building Self-Discovering Analytics with Praval, an open-source Agentic AI framework more
  • 2 comments
  • Submitted
  • 10 Nov 2025
Type of submission: 15 mins talk

Manish Malik

How are we building AI Agents for Fraud Detection?

At Razorpay, we process billions of transactions. This scale brings a massive, evolving threat from fraud—from phishing scams to money laundering. In this talk, we’ll start by showcasing the ML capabilities Razorpay uses to prevent platform abuse by bad actors posing as legitimate merchants. When our models flag a merchant, the real work begins: conducting a thorough risk review. Traditionally, t… more
  • 1 comment
  • Submitted
  • 10 Nov 2025
Type of submission: 30 mins talk

Mayur Madnani

Building Multi-Agent AI Systems: A Philosophical Debate

Abstract This hands-on workshop introduces participants to multi-agent AI systems through a unique philosophical lens. Attendees will learn to build a debate system where three AI agents—embodying Socrates, Plato, and Aristotle—engage in dynamic conversations on user-provided topics. The workshop demonstrates core agentic AI concepts including agent orchestration, persona design, memory managemen… more
  • 0 comments
  • Submitted
  • 10 Nov 2025
Type of submission: Hands-on workshop (1.5 hrs to 3 hrs)

Mayur Madnani

Fine-Tuning SLM models with LoRA: Building Specialized On-Device Story Generators

Context and Background Generative AI has opened up new creative possibilities, but most applications rely on cloud-based models, which can introduce latency, privacy, and cost challenges. In this session, we explore how to shift from server-side story generation to efficient on-device personalization. By using LLM and SLM models, we demonstrate a hybrid approach that maintains quality while optim… more
  • 0 comments
  • Submitted
  • 10 Nov 2025
Type of submission: Hands-on workshop (1.5 hrs to 3 hrs)

Darshan Adiga Senior Data & Applied Scietist

Lakshya Daulani Data & Applied Scietist

Enterprise-ready & Compliant Synthetic Data Generation for Data Governance

In AI driven workflows, a critical bottleneck is the scarcity of realistic training datasets—and strict privacy and security rules forbid using actual customer records as well. To address these challenges, we have developed an agentic synthetic data generation pipeline that produces domain-rich, realistic, and coherent datasets, for training PII (Personally Identifiable Information) and sensitive… more
  • 1 comment
  • Submitted
  • 10 Nov 2025
Type of submission: 15 mins talk

Shwetank Dixit

From SQL to Conversations: Towards a Friendlier Analytics Platform for Merchants

Traditional analytics tools force users to think like data analysts—writing queries, understanding schemas, and interpreting dashboards. Most small sellers on platforms like Shopify are overwhelmed with the sheer amount of data there is. What if analytics could understand human questions instead? more
  • 1 comment
  • Submitted
  • 10 Nov 2025
Type of submission: 30 mins talk

Mudit

[Always on Insights - powering performance of media campaigns]

One of the key challenges in the retail media ecosystem is to personalize stories for advertisers and deliver them with precision and speed. Target’s retail media business, Roundel, addresses this challenge through Premium Insights. At its core, Premium Insights engages and pitches to advertisers by automating the creation of compelling, data-driven media and consumer stories. The solution levera… more
  • 1 comment
  • Submitted
  • 10 Nov 2025
Type of submission: 30 mins talk
Siddhant Agarwal

Siddhant Agarwal

Demystifying the Mahabharata with GraphRAG

“Demystifying the Mahabharata with GraphRAG” takes you on a journey where ancient wisdom meets modern AI innovation. This session delves into how GraphRAG — an emerging approach that fuses knowledge graphs with Large Language Models (LLMs) — can be used to decode the vast narrative universe of the Mahabharata. By structuring the relationships between characters, events, and moral dilemmas in grap… more
  • 0 comments
  • Submitted
  • 11 Nov 2025
Type of submission: 30 mins talk

Prasanna Bhogale

Your Causal Parrot might be lying to you

The talk will start with an introduction / motivation about causal inference in general. Humans have excellent causal models of the world, so associational analytics suffice in human businesses because the causal model of the business is implicitly known to everyone involved, thats why - despite almost no one bothering with causal inference - most BI departments are doing just fine. On the other … more
  • 1 comment
  • Submitted
  • 12 Nov 2025
Type of submission: 30 mins talk
Abhiram Ravikumar

Abhiram Ravikumar

Video thumbnail

Stop Guessing: A Practical Playbook for Measuring Prompt Quality

Theme fit: Semantic layers & AI in practice Overview more
  • 0 comments
  • Submitted
  • 12 Nov 2025
Type of submission: 30 mins talk

krishan goyal

Reimagining Analytics: Unlocking AI-Driven Insights with Pinot MCP Server

Description This session will showcase the potential of Pinot MCP Server—a capability which enables business teams to interact with enterprise data through natural language. more
  • 1 comment
  • Submitted
  • 13 Nov 2025
Type of submission: 30 mins talk

Vikram Nayak

Live Wireframing: Lean BI using AI

About the session: This talk explores dashboard wireframes as lightweight thinking tools that enable rapid iteration before any technical implementation begins. Like Lean Startup’s “build-measure-learn” loop, but for dashboards - wireframe, review, refine, then build. more
  • 0 comments
  • Submitted
  • 14 Nov 2025
Type of submission: 30 mins talk

Nishant Bangarwa

From Data to Decisions: An End-to-End AI + BI-as-Code Workshop

AI is reshaping how teams explore and reason about data. Business users increasingly expect natural-language answers. Data engineers want analytics they can version, test, and deploy like software. AI researchers need structured semantic models that agents can reliably operate on. To serve all three, analytics must evolve into something more semantic, more interactive, and more code-driven. more
  • 0 comments
  • Submitted
  • 14 Nov 2025
Type of submission: Hands-on workshop (1.5 hrs to 3 hrs)

Balamurugan P

[Talk] Hybrid LSTM and LLM Framework for Anomaly Detection in Real-Time Market Data

Brief Markets move fast, which makes anomaly detection difficult. Models based on tree methods become stale and fail to track time-based patterns or understand complex context. This talk presents a hybrid system that combines LSTM networks with large language models to detect anomalies in real time market feeds. LSTM layers capture time-based patterns at a granular level. LLM layers add semantic … more
  • 0 comments
  • Submitted
  • 15 Nov 2025
Type of submission: 15 mins talk

Balamurugan P

[Talk] Hybrid LSTM and LLM Framework for Anomaly Detection in Real-Time Market Data

Brief Markets move fast, which makes anomaly detection difficult. Models based on tree methods become stale and fail to track time-based patterns or understand complex context. This talk presents a hybrid system that combines LSTM networks with large language models to detect anomalies in real time market feeds. LSTM layers capture time-based patterns at a granular level. LLM layers add semantic … more
  • 0 comments
  • Submitted
  • 15 Nov 2025
Type of submission: 15 mins talk

Sushrut Ikhar

Future Proofing ML Inferencing: High Performance Java meets Scalable Remote Serving

In this talk, we present a hybrid ML inference architecture designed for today’s performance demands and tomorrow’s infrastructure challenges. We combine SIMD-accelerated inference in pure Java—using the Vector API and Fused Multiply-Add (FMA)—with remote inference via TensorFlow Serving, ONNX, and Triton. This allows us to strike the right balance: ultra-low latency and high throughput on critic… more
  • 0 comments
  • Submitted
  • 16 Nov 2025
Type of submission: 15 mins talk

Hosted by

The Fifth Elephant
The Fifth Elephant
Jumpstart better data engineering and AI futures