Skip to content
AICDAI Community Days
Event poster for AICD - Mumbai
9 speakersannounced so farLocal teamRun by Jamel, Mourtaza and 2 more

About AICD - Mumbai

🌐 AI Community Days (AICD): A Global Movement for AI Knowledge Sharing

AI Community Days (AICD) is more than a conference it’s a worldwide initiative to democratize Artificial Intelligence and make innovation accessible to everyone.

What Makes AICD Unique?

Community-Driven: Organized by passionate AI professionals, developers, and enthusiasts not corporations.

Global Reach: Events in 30+ major cities across every continent, including San Francisco, Tokyo, Bengaluru, London, Paris, Frankfurt, Dubai, New York, Toronto, São Paulo, Seoul, Sydney, and Riyadh.

Inclusive & Ethical: Focused on responsible AI, diversity, and empowering local communities.

What Happens at AICD Conferences?

Read the full description

Keynotes & Expert Talks: Learn from global AI leaders and innovators.

Hands-On Workshops: Practical sessions on AI tools, frameworks, and real-world applications.

Panel Discussions: Explore ethics, governance, and the future of AI.

Networking Opportunities: Connect with peers, mentors, and industry influencers.

Community Showcase: Highlight local projects and success stories.

Why Join AICD?

Free & Accessible: All events are community-driven and free.

Knowledge Sharing: Gain insights into cutting-edge AI trends.

Global Collaboration: Build partnerships across borders.

Empowerment: Help shape the next generation of AI talent.

The future of AI is collaborative. Let’s build it together.

Agenda

The timetable is on its way

8 sessions are confirmed. Times and rooms are added once the schedule is final. Save the date: 21 November 2026.

See the sessions

All sessions

Times in Mumbai time

  • AI Fluency for Leaders: From Knowing AI to Leading with AI

    AI literacy is becoming increasingly common. But for leaders, knowing what AI can do is only the beginning. As AI becomes embedded into everyday work, leaders need a deeper capability: **AI fluency**—the ability to recognize opportunities, ask better questions, make informed decisions, understand risks and lead people through AI-driven change. This session explores the difference between **AI awareness, AI literacy and AI fluency**, and why the distinction matters for modern leadership. Participants will explore five dimensions of AI-fluent leadership: understanding AI possibilities, identifying business opportunities, leading responsible adoption, redesigning work and enabling people to adapt. We’ll also examine how leaders can balance experimentation with governance, speed with responsibility, and AI-driven efficiency with human-centered leadership. The session provides a practical roadmap for leaders who don't need to become AI engineers—but do need to become confident leaders in an AI-enabled workplace.

  • From AI Prototype to Production: Lessons from Deploying Video AI at Scale

    AI demos are impressive. But taking an AI application into production is where the real engineering challenge begins. Drawing from real-world lessons learned while deploying VideoAiQ, an AI-powered video intelligence platform, on Jio Azure, this session explores the journey from AI experimentation to a scalable and resilient production deployment. The session will uncover practical challenges around cloud infrastructure, compute and GPU planning, video processing at scale, inference latency, Kubernetes, observability, reliability, security, and cloud cost optimization. It will also highlight how traditional DevOps and SRE practices need to evolve to address the unique operational characteristics of AI workloads. Rather than focusing on AI models and algorithms, this talk presents an infrastructure and production engineering perspective on AI—sharing architectural decisions, deployment challenges, and key lessons from the field. Because building an AI model is only the beginning. The real test is making AI scalable, observable, cost-efficient, and reliable in production.

  • From Developer to AI Engineer: Skills Every Software Professional Needs by 2027

    The software industry is shifting toward AI-native development, creating demand for engineers who understand both traditional application architecture and intelligent systems. This session provides a practical roadmap for developers transitioning into AI engineering roles. We will explore retrieval-augmented generation, vector databases, model evaluation, prompt engineering, agentic workflows, and secure AI deployment practices. Participants will gain insight into emerging technologies and discover the skills required to design, deploy, and maintain intelligent applications that deliver measurable business outcomes.

  • Mastering Copilot Studio governance

    In this session we talk about efficient governance of copilot studio agents. We talk about not only the settings and configuration but also discuss best practices and use cases.

  • Meet Microsoft Scout: Building Always-On AI Agents

    Discover Microsoft's latest vision for proactive AI agents that continuously monitor events, automate workflows, and take intelligent actions without constant user prompts. Explore practical enterprise use cases and future possibilities.

  • Reimagining Productivity with Microsoft AI

    Artificial Intelligence is reshaping how we work. Discover how Microsoft AI and Microsoft 365 Copilot can automate repetitive tasks, enhance collaboration, generate high-quality content, and streamline everyday workflows. Through practical demonstrations and real-world scenarios, this session showcases how individuals and organizations can boost productivity while adopting AI responsibly.

  • Turning Probabilistic Agent Trajectories into Production Skill Stores

    Deploying autonomous agents to handle multi-step operational workflows (e.g., incident triage, cloud resource migration, database re-indexing) faces fundamental production bottlenecks: probabilistic latency and variance to name a few. When an LLM agent executes a 10-step workflow, every step involves a probabilistic tool call. This creates three critical issues in production environments: 1. Compounding Non-Determinism: A 95% single-step accuracy rate degrades to a ~60% success rate across a 10-step agent loop. 2. Latency & Cost Inflation: Re-evaluating system prompts, tool schemas, and multi-turn history on every single request wastes millions of tokens and introduces severe execution delays (often taking 30–90 seconds per task). 3. Context Rot & Drift: As intermediate tool outputs accumulate in the context window, the agent risks missing edge-case constraints or sometimes gets stuck in retry loops. Our Approach: Instead of forcing LLMs to re-evaluate repetitive multi-step plans, our Agent Compiler converts dynamic agent trajectories into deterministic code. When an agent solves a new problem, an offline node sanitizes its execution logs, injects schema guardrails, and compiles the sequence into a statically typed script saved to a centralized Skill Store. Future matching requests bypass the LLM planning loop via semantic dispatch—running the compiled artifact directly at 1/100th the token cost and 10x the speed. Tech-Stack - Agent Orchestration: LangGraph / FastAgency / OpenTelemetry Tracing - Compiler & Parsing Layer: Python (AST, Pydantic v2 runtime validation) - Tool Protocol: Model Context Protocol (MCP) - Skill Store Infra: Redis / VectorDB + Git/S3 for script versioning Target Audience: Software Engineers, Architects, Applied Scientists, Solution Engineers, Problem Solvers. Key Takeaways: 1. Master the architecture for capturing dynamic, probabilistic LLM reasoning and compiling it into fast, deterministic, statically typed code. 2. Learn exact engineering strategies to slash token costs by up to 90% and cut multi-step task execution latency from 45 seconds down to under 1 second. 3. Design a version-controlled, rollback-ready Skill Store with automated contract testing, semantic dispatch routing, and human-in-the-loop review boundaries.

  • Your AI Agent Has Access to Production. How Do You Build Trust? — Securing Agentic AI

    AI is rapidly evolving from chatbots that simply generate responses to autonomous agents that can access APIs, databases, files, external tools, and even perform actions on behalf of users. But with this increased capability comes a new security challenge: What happens when an AI agent is manipulated into doing something it shouldn't? In this session, we will explore the security risks that emerge when building real-world agentic AI systems, including prompt injection, indirect prompt injection, excessive permissions, tool misuse, sensitive data exposure, and agent identity and authorization. Through architecture examples and a practical demo, we will see how an apparently helpful AI agent can be manipulated into performing unintended actions — and, more importantly, how we can design the system to prevent them. We will then look at practical security patterns such as least-privilege tool access, authentication and authorization, policy enforcement, input/output validation, human approval for sensitive operations, and monitoring of agent actions. The goal is simple: don't just build AI agents that work — build AI agents you can trust.

Speakers

9 speakers announced so far.

Open the wall

The team behind Mumbai

The local team organizing AICD - Mumbai.

Open the wall

Global team

The people who run AI Community Days around the world.

Be part of it

Help make Mumbai happen.

Every AI Community Day is run by volunteers from its city. Here is how you can help this one.

  • Sponsor Mumbai

    Put your brand in front of the people building with AI in Mumbai. Ask for the sponsor package.

    Sponsor this event

  • Volunteer at the event

    Help with check-in, rooms and social media, and meet everyone behind the scenes.

    Volunteer

Upcoming events

Can't make it to Mumbai?

Each event is run by a local team.

All events and cities