The Fifth Elephant 2026 Annual Conference
Built for humans. Now rebuilding for agents.
Accepting submissions
Not accepting submissions
Nilesh Mahajan Presenter Breaking the Trilemma: Serverless Data Platform in Your Own Cloud AccountDescription If you run data workloads, you’ve probably hit the same wall we did. You want compute that stays in your own cloud account, runs fast, and stays cheap — and sooner or later, someone tells you to pick two. This is a hard problem we solved for our own platform, and this talk shares those learnings around the engineering it takes to get all three. more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
|
AC
Alex Campos Workshop facilitator Streaming Meets the Lakehouse: Hands-on with Fluss and IcebergAbstract: In this hands-on workshop, attendees will build a Streaming Lakehouse using Apache Fluss and Apache Iceberg. Learn to ingest real-time streams, store them in Fluss tables, and transparently tier data into Iceberg for long-term analytics. Through guided exercises, you’ll define Flink SQL pipelines, enable Fluss tiering to Iceberg, and run real-time and historical queries. By the end, you… more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: Hands-on workshop - 2-4 hours
|
Dhruv Nigam  Your voice agent is (probably) doomed, OR how not to fall victim to outdated voice agent playbooksGoogly Bhai was busy this IPL season. He live-streamed to an audience of 300,000 every day, with 1.5 million minutes of watch time and 1.1 million concurrent viewers at peak. Hundreds of other streamers called him onto their streams to discuss live scores, gossip, and make predictions (see him live jamming with another streamer). He switched to Aussie and British accents mid-conversation at the a… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 30 mins talk
|
|
VK
Vivek Kalyanarangan Senior Technical Architect - AI Tech at IDfy Beyond GPUs: Cutting ML Inference Costs by 10× Without Sacrificing LatencyInference cost-to-serve is usually treated as a fixed tax: the model needs a GPU, the GPU costs what it costs, and the bill scales with traffic. It isn’t fixed. For a large class of production models — embeddings, CNNs, classic CV and NLP — quantization plus graph fusion turns that GPU tax into a variable you control, cutting cost-to-serve by ~10× at the same latency, throughput, and accuracy env… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 30 mins talk
|
Shivam Gupta Discover Globally, Materialize Locally: Building a Governed Cross-Domain Data Sharing PlatformCross-business-unit data sharing usually starts with good intentions and quickly turns into ticket-driven exports, undocumented copies, governance bottlenecks, and growing compliance risk. At InMobi, multiple business units operate independent lakehouses, catalogs, and data engineering organizations. Combining data across these domains creates significant business value, but traditional approache… more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
Sujeet Gholap Architecting Observability Platform on S3 + LambdaObservability is a Data Engineering problem. Write-heavy, realtime latency, faster queries are the typical requirements of a Observability system. Traditional observability systems fail to meet the ever-growing demand of increased volumes due to cloud deployments, AI agents speeding up the feature and product development. AI agents also changes the query patterns which were common in Observabilit… more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
Harshad Nawathe The Practical Guide to Reverse-Engineering XXL Codebases with Agentic AIAbstract Point an AI coding agent at an unfamiliar codebase and ask it to explain the architecture, and for a few hundred files, it works beautifully. Point the same agent - with the same well-engineered prompt - at a ten-thousand-file enterprise Java monolith, and it quietly falls apart: the context window fills, compaction kicks in, the model starts forgetting details, and then it starts invent… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 30 mins talk
|
Sathish Presenter AIOps: Leveraging AI for Software IncidentsDescription During production incidents or on-call schedules with a barrage of alerts, engineers must sift through hundreds or thousands of services, code changes, metrics data points, logs, and traces to reason about the issue and find the root cause. more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 30 mins talk
|
Kalpesh Jajoo Founding Solutions Architect, APAC at Skyflow Don’t Block AI: Handling Sensitive Data and DPDP While Preserving ContextDescription Enterprise AI is forcing organizations to rethink one of the most fundamental assumptions in data security. For decades, security strategies have focused on protecting data where it is stored. Encryption, tokenization, access controls, and governance were designed for applications that queried databases and presented information to users. Generative AI changes that model completely. T… more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
|
AN
Abhijith Neerkaje Workshop instructor AI evals workshopOverview Why do Agents make mistakes - 3 Gulfs [Comprehension, Specification and Generalization]. (10 min) more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: Hands-on workshop - 2-4 hours
|
Rajath 
Lakshmi Narayana G From Files to CatalogsWorkshop: From Files to Catalogs — Modern Data Foundations with Parquet, Iceberg, and Polaris more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: Hands-on workshop - 2-4 hours
|
Amit Prabhu Full Refresh to Incremental: Rebuilding Denormalization for Reporting at Razorpay ScaleAbstract / Session Description: At Razorpay, our lakehouse platform ingests over 6 billion events daily and powers a reporting platform that generates close to a million reports every month. As scale grew, full-refresh denormalization became unsustainable: joins across 10-30 entities consumed heavy compute, report freshness lagged by up to 48 hours, and highly mutating datasets made derived table… more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
Karthika Vijayan When the Orchestrator Wasn’t Enough: A Production Story of Building Reliable Multi-Agent AIWe built a production multi-agent assistant for a complex enterprise planning workflow where AI assists, rather than replaces, human decision-making. Users express their intent through natural language, the system extracts structured semantic information for human review, and multiple specialist agents collaborate to retrieve data, perform analysis, and generate recommendations. Our initial archi… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 30 mins talk
|
Vikram Nayak When There's No Unit Test for "Good": A Maker-Checker Loop for Subjective AI OutputBRIEF DESCRIPTION: The problem: When an AI agent writes code, you can test whether the code works. But when an agent makes a chart, how do you test whether it’s any good? A chart can be technically correct and still fail to get its point across. “Good” depends on the audience and the decision they need to make - there’s no test that returns true or false. more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 30 mins talk
|
Pravein Govindan Kannan Optimized AI Inference with llm-d and a case-study on Sovereign AI on heterogenous GPU clusterIntroducing llm-d llm-d (llm-d.ai) is a high-performance distributed inference serving stack optimized for production deployments on Kubernetes It provides a transparent routing layer that sits between the client and vLLM serving pods, making key scheduling decisions at request granularity: more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
Vivek Sinha High Concurrency & Low Latency Serving on Apache IcebergDescription Everyone is putting their data into Apache Iceberg; almost no one is serving sub-second queries directly from it. Once data lands in Iceberg, a familiar question arises: how do you power real-time experiences without duplicating it in yet another serving system? This challenge is especially sharp in observability workloads like RUM, clickstream, and APM, and in customer-facing analyti… more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
Mayur Singal Beyond Metadata: Building an Open Context Layer for AIDescribe your session in 2 paragraphs Everyone has seen how AI transformed software engineering. Tools like Claude Code and Codex work because code comes with rich context—source control, dependency graphs, reviews, tests, and execution history. Enterprise data, however, lacks an equivalent foundation. AI agents are expected to answer questions, generate SQL, understand business metrics, and auto… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 30 mins talk
|
Mayur Jadhav An Agent That Builds Agents: AI-Powered Recipe Generation for Local-First ETLAn Agent That Builds Agents: AI-Powered Recipe Generation for Local-First ETL more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 30 mins talk
|
Amaan Shaikh Billboards near a coffee shop: making proximity search fast in ClickHouseTitle Billboards near a coffee shop: making proximity search fast in ClickHouse more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
Rajmohan C Diagnosing data pipeline failures with AI agents: from research to production, and the open challengesModern enterprise data platforms rely on complex data pipelines for data transformation and integration, with thousands running every day to move data across systems. When a data pipeline run fails, the error you see is usually not the root cause. Failures surface far from where they are born, and the component that throws the error is rarely the one at fault; so diagnosing them means reasoning a… more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
|
SR
Shashank Rao Facilitator Beyond SWE-bench: How do we evaluate AI agents for real-world workflows?Track: Track 2 – Building & Implementing AI Tools & Agents in Production Format: Birds of a Feather (BoF) Session more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: Birds of Feather (BOF) session
|
Prateek Mandloi 
Sunny R Gupta How to Engineer Agent Quality Loops (White-Box) for Reliable AI Agents?Description Agent systems work through loops: plan, call tools, inspect results, recover, and decide when to stop. As these systems move beyond demos, the harder question is not whether they can complete a task, but whether the path was reliable, efficient, grounded, and worth its cost and what should change when it was not. more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: Birds of Feather (BOF) session
|
Priyanga P Kini Designing Reliable Agentic LoopsOverview “Loops” are becoming a popular mental model for how we build with agents. Agents reach their goals by planning, calling tools, inspecting results, retrying failed steps, using memory, and sometimes calling other agents. As these loops become more autonomous, the question is no longer just whether the agent produced a good final answer, but whether each part of the loop is working reliabl… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: Birds of Feather (BOF) session
|
Pranav Bhosale Vector DBs Are Overrated: Grounding LLM Agents on Master Data Without the OverheadLarge Language Models are powerful at reasoning and recommendation, but they routinely hallucinate entities that do not exist in an organization’s proprietary data. The common response is to build a Retrieval-Augmented Generation (RAG) stack with embeddings, a vector database, and retrieval infrastructure. However, many enterprise datasets are not collections of long documents—they are structured… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 15 mins talk
|
Omkar Sabade From Figma Node to Production Code: How we Taught Agents to build UI in PhonePeEvery design system has the same expensive seam: the handoff from Figma to code. A designer ships a pixel-perfect screen, and an engineer spends hours inspecting layout, colors, spacing, and variants, then hand-writing it into the codebase over and over, across every platform the design system targets. At PhonePe we built FigGen, an agentic pipeline that turns a selected Figma node into productio… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 15 mins talk
|
Navdeep Agarwal From PDF to SQLFifth Elephant 2026 Submission Description Most “ask my documents” systems stop at retrieval: upload PDFs, create embeddings, and search for relevant chunks. That works when the user wants surrounding context, but it starts breaking down when the user needs numbers for a report, dashboard, analysis, or presentation. In Indian mutual fund reports, the useful answers are usually structured facts: s… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: 15 mins talk
|
Hrithik Piyush Many Narrow Agents, Not One Big Prompt: Putting AI on the On-Call RotationAbstract The obvious way to put an AI agent on the on-call rotation is one large prompt that knows everything. We rejected it for a multi-agent architecture: around 18 specialized agents over a large cloud database, a router that classifies each incident to the right specialist, and a shared library of nearly 300 reusable skills. Diagnosis is automated but mitigation stays human-approved. more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 15 mins talk
|
Swetha A 
Mahita D  From Amnesia to Photographic memory in Agentic SystemsObjective / Abstract : You’ve built a few agents, wired them together into a multi-agent system. What comes next? This workshop takes a focused look at one of the most critical and often underestimated parts of an agentic system: memory. Participants will come away with a practical understanding of how memory shapes agent behavior, capabilities, and user experience. more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: Hands-on workshop - 2-4 hours
|
Your Database Isn’t Ready for Agents—YetThe misconception is that an “AI-ready” database only needs a vector index and native RAG, which leads to state fragmentation: juggling object storage, a relational database for metadata, and a vector database. more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
|
SK
Sujit Kamthe Birds of Feather facilitator Things We Wish We Knew Before Building Agentic SystemsEveryone is building agents. Most demos look impressive. Production systems tell a different story. more
|
|
Karthika Vijayan BOF facilitator Agentic AI: Path to Production - A Birds of a Feather DiscussionAbstract Building an agent that works in a demo is no longer the hard part. The challenge begins when the same system must operate reliably in production, handling evolving requirements, multiple agents, long-running workflows, changing prompts, evaluation, latency constraints, and operational failures. Every team encounters different problems, yet many of the underlying engineering questions are… more
|
Akash Sathish Preserving Human Intuition in AI DevelopmentThree major studies from Anthropic, METR, and MIT recently revealed an uncomfortable reality about AI-assisted development. Their data shows that AI tools can lower skill assessment scores, increase actual task completion times, and significantly reduce neural connectivity during creative problem-solving. This is not a simple case of developers becoming lazy. Instead, widespread AI usage actively… more
I am submitting for: Track 2 - Building & implementing AI tools & agents in production
Type of session: Birds of Feather (BOF) session
|
Why Your Database Architecture Will Make or Break Your AgentAs AI agents become more autonomous, databases are emerging as one of the biggest bottlenecks to scale. This session explores how AI-native applications fundamentally change database requirements, from handling massive write throughput and parallel agent execution to supporting rapid iteration and resilient growth. Through the lens of the Manus AI journey, we’ll discuss the architectural principl… more
I am submitting for: Track 1 - Data engineering & infrastructure
Type of session: 30 mins talk
|
Get your hybrid access ticket
Hosted by
Jumpstart better data engineering and AI futures
Supported by
Platinum Sponsor
Atlassian unleashes the potential of every team. Our agile & DevOps, IT service management and work management software helps teams organize, discuss, and compl
Platinum Sponsor
Sahaj is an artisanal technology services company crafting purpose-built AI and data-led solutions for businesses.
