BEGIN:VCALENDAR VERSION:2.0 PRODID:-//HasGeek//NONSGML Funnel//EN DESCRIPTION:Share you work with the community X-WR-CALDESC:Share you work with the community NAME:Speak at The Fifth Elephant 2026 Annual Conference X-WR-CALNAME:Speak at The Fifth Elephant 2026 Annual Conference REFRESH-INTERVAL;VALUE=DURATION:PT12H SUMMARY:Speak at The Fifth Elephant 2026 Annual Conference TIMEZONE-ID:Asia/Kolkata X-PUBLISHED-TTL:PT12H X-WR-TIMEZONE:Asia/Kolkata BEGIN:VEVENT SUMMARY:Speak at The Fifth Elephant 2026 Annual Conference DTSTART:20260731T033000Z DTEND:20260731T123000Z DTSTAMP:20260709T192744Z UID:session/81v9ryXuyQTNgK5uMxoJyz@hasgeek.com SEQUENCE:22 CREATED:20260529T065954Z DESCRIPTION:# This page is only for submissions for The Fifth Elephant 202 6 annual conference\n**Dates** \n- Conference on Friday\, 31 July\, at the NIMHANS Convention Centre \n- Workshops in July (7\, 17\, 25) and Saturda y\, 1 August in Marathalli\, Koramangala & Whitefield\n\n**To attend the c onference\, get an annual membership - https://hasgeek.com/fifthelephant#m emberships**\n\n---\n\n# How to submit\n1. **Submit your abstract**. Your abstract should clearly describe:\n * the problem or topic being address ed\,\n * why it is relevant\,\n * who the session is intended for\, an d\n * the key takeaways attendees can expect.\n\n2. **Add a link to draf t slides within 3ā€“5 days** - after submitting your abstract\, add a link to the draft version of your slides within 3ā€“5 days. Draft slides help editors better evaluate the structure\, depth\, and delivery of the sessio n.\n*Submissions without draft slides may not be reviewed or may experienc e delays in receiving feedback.*\n\n3. **Add a 2-minute elevator pitch vid eo** - introducing your session and demonstrating your speaking style. Thi s helps the editorial team assess presentation clarity and audience engage ment.\n\n4. **Make your submission here - https://hasgeek.com/fifthelephan t/fifthelephant-2026-call-for-submissions/sub**\n\n---\n\n# Review timelin es and submission deadlines\n* Feedback via comments may be shared early f or submissions that include slide links.\n* Selected speakers will be cont acted starting 17 June onwards.\n* The final submission deadline is 25 Jun e 2026.\n* Not every submission may fit the annual conference schedule in July. Some proposals may be considered for future community sessions\, inc luding weekly reviews and monthly meet-ups.\n\n--- \n\n# Conference trac ks\n# Track 1: Data Engineering & Infrastructure\n\n*The hard problems are just beginning.*\n\nThe orchestration tools have matured. The primitives are no longer the argument.\n\nWhen agentic systems generate ad hoc querie s that shatter your partition assumptions\, when a pipeline written by an AI model goes to production with your name on the approval\, when an auton omous quality monitor acts on a data quality miss before a human sees it\, and when your cloud bill arrives with a line item no one planned for.\n\n **This track is built around a single mandate: real systems under real pro duction constraints : latency\, cost\, and scale.** Track editors are look ing for submissions that reflect what's actually working (and what isn't) in production data systems\, especially as AI workloads reshape infrastruc ture assumptions.\n\nWe want the story of what broke\, what you threw away \, and what you rebuilt to survive.\n\n---\n\n## AI-native data systems\n\ nWho is consuming your data\; humans or agents? When it's agents\, the eng ineering decisions change fundamentally. Token budget replaces query cost. Descriptions of relationships matter more than the relationships themselv es. Functions beat tables. Logs beat docs.\n\nWe want talks on:\n\n- Data engineering for foundational model training and fine-tuning\n- Data pipeli nes for inference and post-training steps\n- Data stores and retrieval pat terns for GenAI and LLM-based applications\n- Memory management and agent- to-agent communication\n- Evals\, guardrails\, and observability implement ed in production systems\n\n---\n\n## Foundations & storage\n\nThe format wars produced two survivors worth betting on. The catalog and query engine landscape did not stand still. And the assumption that you need a distrib uted cluster for serious workloads is being quietly dismantled.\n\nWe want talks on:\n\n- Lakehouse and lakebase architectures: what's working in pr oduction\n- Table format evolution: Iceberg\, DuckLake\, and how metadata management is changing\n- Query engines in the wild: including newer entra nts like Apache DataFusion\n- In-memory and local databases: when and why they make sense\n- Realtime CDC and streaming data patterns\n\n---\n\n## G overnance\, compliance & data quality\n\nGovernance frameworks that exist only in documentation are not what we are looking for. The metadata layer has moved from nice-to-have to load-bearing infrastructure\; and agents ha llucinate metrics without a governed definition layer. Meanwhile\, DPDP is no longer a future problem.\n\nWe want talks on:\n\n- Practical approache s to PII detection\, masking\, and management at scale\n- Metadata managem ent strategies that actually hold up in large organisations\n- Governance frameworks adapted for AI and LLM pipelines\n- Data quality practices in t he context of model inputs and outputs\n\n---\n\n## Ops\, reliability & co sts\n\nCloud bills are bleeding\, and the GPU line item has arrived. Embed ding and vector compute are now cost centres that data engineers own\, not ML platform teams. And in closed-loop systems where agents act on data be fore a human sees it\, standard observability pillars are no longer enough .\n\nWe want talks on:\n\n- Observability practices for data and AI infras tructure\n- Agent incident response and SRE patterns for agentic systems\n - Cost optimisation strategies for AI infrastructure: compute\, storage\, egress\n- Hardware and compute considerations\; sub-architecture patterns for inference and training workloads\n\n---\n\n## Orchestration & pipeline s\n\nThe engineer who gave an agent write access to production has a diffe rent relationship with orchestration than the one who runs nightly batch j obs. Migration stories\, operational failure modes\, and the integration o f agent workflows into pipeline tooling.\n\nWe want talks on:\n\n- Product ion use cases with tools like Airbyte\, Fivetran\, Dagster\, Prefect\, or Temporal\n- Patterns for integrating orchestration with AI and agent workf lows\n- Migration stories from legacy pipelines to modern orchestration\n- Operational lessons from running pipelines at scale\n\n---\n\n## A standi ng invitation to disagree\n\nWe are holding a slot for the talk that pushe s back on everything above.\n\nMost companies are not running agentic data systems. Most data engineering is still ETL\, warehousing\, and batch pro cessing\; unglamorous\, unfinished\, and real. If your strongest opinion i s *"your team doesn't need agents\, it needs a working warehouse and teste d pipelines"*. You can back it with production experience. It keeps the re st of the programme honest.\n\n---\n\n# Track 2: Building and Implementing AI Tools & Agents in Production\n\n*The demo always works. Production is where every assumption breaks.*\n\nEvery team has now built an agent. Most of them work brilliantly in the demo. A smaller number are actually in pr oduction. A smaller number still are in production and behaving as designe d six months later.\n\nThe gap between those groups is not a model problem or a prompt problem. It is testing reliability under real load\, observab ility when the system makes a decision you didn't anticipate\, cost when t he token bill arrives\, and accountability when an agent acts on bad input before a human notices.\n\n**This track is for the engineers and teams wh o have crossed into production\, or are close enough to see what's waiting for them there.** We want the case study where the architecture changed a fter the first real incident. The integration that took three attempts to get right. The evaluation strategy that replaced gut feel. The business ca se that survived contact with a CFO\n\n---\n\n## Building production-ready agents and tools\n\nBuilding one that holds up under real usage\, degrade s gracefully\, and doesn't surprise you at 2am is a different discipline e ntirely. The MCP ecosystem has matured enough to have opinions about\; and enough production scar tissue to share.\n\nWe want talks on:\n\n- Buildin g production-ready agent frameworks and tools\n- Implementing MCP (Model C ontext Protocol) servers and clients\n- Tool development for specific doma ins: customer support\, operations\, data analysis\n- Designing tool inter faces for LLM consumption\n- Creating reusable agent components and librar ies\n\n---\n\n## Multi-agent orchestration\n\nWhen one agent calls another \, you have distributed systems problems again\; latency\, partial failure \, state management\, and the unique joy of debugging a chain where every step was technically successful and the output is still wrong. Orchestrati on at this layer is genuinely unsolved\, and practitioners are working it out in production.\n\nWe want talks on:\n\n- Multi-agent orchestration pat terns that held up under real conditions\n- Scaling agent-based systems: w here the architecture broke and replacements have to be done\n- Reliabilit y and monitoring for agentic systems\n- Operational challenges and solutio ns from running agents at scale\n\n---\n\n## Agents in the enterprise\n\nE nterprise environments bring SSO\, procurement\, compliance reviews\, chan ge management\, and colleagues who did not ask for an AI agent in their wo rkflow. The integration stories from teams who navigated all of that are t he ones this audience needs.\n\nWe want talks on:\n\n- Real-world case stu dies of agents in production: what changed between v1 and what runs today\ n- Integration patterns for enterprise environments\n- Migrating from trad itional automation to AI agents\; the technical and organisational frictio n\n- ROI and business impact of agentic systems subject to what can be sha red.\n- Team workflows with AI agents\; what changed for the people\n\n--- \n\n## Evaluation\, testing & quality assurance\n\nHow do you write a test for a system that is non-deterministic by design? How do you know your ag ent is getting better when "better" is partly subjective? These are genuin ely hard problems\, and the field is developing real answers.\n\nWe want t alks on:\n\n- Agent evaluation strategies and quality assurance frameworks \n- Testing strategies for non-deterministic systems\n- Prompt engineering and agent optimisation\; what actually moved the metric\n- Debugging appr oaches when the failure mode is "the reasoning was plausible but wrong"\n\ n---\n\n## Observability\, security & human-in-the-loop\n\nWhen an agent t akes an action in a closed loop\, the standard observability stack tells y ou what happened. Security and human-oversight patterns are not add-ons to agent architecture\; they determine whether the system is trustworthy eno ugh to run unsupervised.\n\nWe want talks on:\n\n- Observability and debug ging for agentic systems in production\n- Security and safety in agentic s ystems: enforcing limits that hold under adversarial conditions\n- Human-i n-the-loop patterns: where they are genuinely necessary and where they bec ome a bottleneck\n- Cost optimisation strategies: token budgets\, caching\ , batching\, and the real bill\n\n---\n\n## Tool and framework selection\n \nThe framework landscape changes faster than most teams can evaluate it. The engineers who have actually run LangChain\, LlamaIndex\, CrewAI\, Temp oral\, or a bespoke stack through a production incident have something to say that a benchmark table cannot. Honest trade-off analysis from someone who made the call and lived with it is the most useful thing in this space right now.\n\nWe want talks on:\n\n- Tool choice and framework selection: how you decided\, what you learned\, what you'd change\n- Migration stori es: between frameworks\, between approaches\, or off a framework entirely\ n- Build vs. buy decisions with real context about what tipped the balance \n\n---\n\n## A standing invitation to disagree\n\nWe are holding a slot f or the talk that pushes back on everything above.\n\nMost agent projects a re still in pilot. Many will not make it to production. The honest talk ab out why poorly defined scope\, unmeasurable success criteria\, organisatio nal readiness that was never there\; is as valuable as any success story. If you shut down an agent project and learned something the field needs to hear\, this stage is yours.\n\n---\n\n# Workshop topics\nTopics include\, but are not limited to:\n* MCP server creation with different protocol va riations\n* Agent frameworks like LangChain and LlamaIndex - which are wid ely used\n* Implementing specific agent patterns (e.g.\, ReAct\, chain-of- thought)\n* Tool development tutorials for agentic systems\n* Integration exercises with real APIs and services\n* Hands-on prompt engineering and o ptimization techniques\n\n---\n\n# Conference editors\n* Jagadish K. (Tryf t)\n* Ramkrishna Reddy Yekulla (Red Hat)\n* Ranganadh Thata (Mico)\n* Sita ram Shelke (NVIDIA)\n* Yash Gandhi (OrcaSheets)\n\n# Consulting editor\n* Ravi Balgi (Datanimbus)\n\n---\n\n## Got a question? Need help?\nšŸ“ž Call or text The Fifth Elephant at (91) 7676332020\nšŸ“§ Email \nšŸ’° For sponsorship inquiries\, email LAST-MODIFIED:20260707T082351Z LOCATION:NIMHANS Convention Centre\, Bangalore - https://hasgeek.com/fifth elephant/fifthelephant-2026-call-for-submissions/ ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/fifthelephant-2026-call-for-submissi ons/ BEGIN:VALARM ACTION:display DESCRIPTION:Speak at The Fifth Elephant 2026 Annual Conference in 5 minute s TRIGGER:-PT5M END:VALARM END:VEVENT END:VCALENDAR