🗓 Full schedule is live → https://hasgeek.com/fifthelephant/2026/schedule

Who is your stack really built for now?

Built for humans. Now rebuilding for agents. The Fifth Elephant 2026 Annual Conference

For years, data platforms, voice interfaces, inference pipelines, and testing frameworks were built for human assumptions — human-readable schemas, human-paced conversations, human-reviewed test cases. Agents are quietly breaking every one of those assumptions.

The Fifth Elephant 2026 brings together practitioners who have already hit these walls in production — at Dream11, IDfy, Onehouse — and rebuilt from first principles, with the numbers to prove it.

Two tracks, one question: what does your stack look like when agents, not humans, are the primary users of it? The Fifth Elephant 2026 brings stories from production and the trenches.

📅 31 July 2026 | 📍 NIMHANS Convention Centre, Bengaluru

⏰ One ticket. Five weeks of workshops + the main conference. Price goes up every Friday until 8 August.

Get your ticket → https://hasgeek.com/fifthelephant/2026/#tickets

Includes all pre-conference workshops (7–25 July, plus 8 August) and the main conference day (31 July).

🎤 Keynote

Rethinking data architecture for the agent era by Sunny Bains (PingCap)

The comfortable assumption is that an “AI-ready” database just needs a vector index bolted onto native RAG. In practice, that leads to state fragmentation — teams end up juggling object storage, a relational database for metadata, and a separate vector database, none of which were designed to talk to each other at agent scale.

Modern agentic systems, with potentially millions of agents acting concurrently, demand convergence, not fragmentation. These agents drive bursty, schema-exploding workloads — creating, dropping, and querying millions of tables, spinning up temporary indexes, and requiring instant, consistent shared state. Legacy database architectures weren’t built for this, and the traditional catalog quickly becomes the bottleneck: metadata operations become the new scalability problem.

This keynote lays out the architectural breakthroughs enabling Agentic Scale — and why next-generation databases have to be agent-first, not human-first with an AI feature added on.

Workshops

🔴 24 July — at Nutanix
Stop guessing. Start measuring.
Hands-on workshop on Agentic Evals.

🔴 25 July — at Nutanix
From Files to Catalogs
Modern data foundations with Parquet, Iceberg, and Polaris.

🔴 25 July — at Nutanix
Streaming Meets the Lakehouse
Hands-on workshop with Fluss and Iceberg. Instructor Alex Campos teaches remotely, with volunteers on the ground to guide participants.

🔴 8 August — at Sahaj Software
From Amnesia to Photographic Memory: Building Memory into Agentic Systems
A hands-on workshop on designing short-term and long-term memory for reliable, personalized AI agents.

🔴 8 August — at Sahaj Software
The AI Engineer’s Playbook: Turning Static Models into Dynamic Agents
A hands-on guide to embeddings, RAG, tool calling, and agentic reasoning — with working code.

Conference tracks

Track 1: Data Engineering & Infrastructure

Data platforms were built for humans. This assumption is breaking as agents read messy data differently and increasingly operate the platform rather than just query it. This track brings the rebuild in progress: Onehouse’s serverless compute breaking the cost/performance trilemma, InMobi’s governed cross-domain data sharing, Razorpay’s shift from full-refresh to incremental at billion-event scale, IBM Research’s sovereign AI inference on heterogeneous GPUs, and fast-scaling infra work from e6data, Oodle AI, StarTree, OrcaSheets and Skyflow.

— Curated by Piyush Goel (Razorpay), Ravi Balgi (Datanimbus), and Ranganadh Thata (Mico)

Track 2: Agentic AI - Platforms and Use Cases

Building for agents in production breaks assumptions engineers didn’t know they were making — that voice is chatbots-with-audio, that GPU cost is fixed, that testing an agent works like testing regular software. This track brings the production numbers behind those corrections: Dream11’s voice agent rebuild for 5M+ daily users, IDfy’s 10× inference cost cuts, Atlassian’s AIOps for incident response, and multi-agent, evals, and orchestration lessons from Sahaj Software, Collate, and Chartboss.

— Curated by Jagadish Kasi (Tryft), Ramakrishna Reddy Y (Red Hat) and Yash Gandhi (OrcaSheets)

Who should attend, and why

• Data engineers and platform teams rethinking how data is structured, stored, and served — now that agents, not analysts, are often the ones querying it
• ML and AI engineers building or scaling agents, voice interfaces, or inference pipelines in production — and hitting walls the standard playbooks don’t cover
• Engineering leaders and architects making infrastructure and cost decisions that will age badly if they’re built on human-era assumptions
• QA, evals, and product folks figuring out how to test and measure systems that don’t behave like traditional software

If your team’s stack was built for people and is now quietly serving agents too, this is where the people solving that problem in production are talking about it.

What does your ticket include

Your ticket covers the entire Fifth Elephant 2026 season — not just one day:

What does your ticket include

• ✅ 7 July — When Data Is for Agents (S Anand)
• 🔴 24 July — Stop Guessing. Start Measuring.
• 🔴 25 July — From Files to Catalogs
• 🔴 25 July — Streaming Meets the Lakehouse
• 🔴 31 July — Main conference day, NIMHANS Convention Centre
• 🔴 8 August — From Amnesia to Photographic Memory
• 🔴 8 August — The AI Engineer’s Playbook
• 🔴 19 September — Iceberg Mini Conference (Bangalore and virtual)

🗓 Full schedule →

Volunteer with us

The Fifth Elephant runs on more than speakers and sponsors — it runs on volunteers. We have 20 roles open across the event, from floor operations to speaker coordination, with tiered deadlines depending on the role.

See open volunteer roles →

Attend

🎟 Conference tickets and annual memberships

Conference ticket

• Cancel any time up to 24 hours before the event for a full refund, no questions asked — this doesn’t apply once you’ve attended any workshop under the ticket (7, 17, or 25 July); attending a workshop locks the ticket.
• Includes access to all Fifth Elephant activities — workshops, meetups, and conferences — for one year.
• Easier to get reimbursed through company/HR.

Annual membership

• Access to all Fifth Elephant activities — workshops, meetups, and conferences — for one year.
• No cancellation.

Sponsorship inquiries

Write to sales@hasgeek.com

Contact

📞 Call or WhatsApp: +91 76763 32020
📧 info@hasgeek.com