Joinal Ahmed

Joinal Ahmed

Senior Architect - AI

About

Joinal Ahmed is an accomplished AI & Cloud Architect with experience working across AWS, Google Cloud and Microsoft, and highly recognized technical speaker with extensive contributions across Cloud Native, MLOps, and Generative AI ecosystems. His expertise bridges theoretical AI concepts with practical, production-ready implementations in scalable and secure enterprise environments. Recognised as Top 3% Most Active Speakers globally on Sessionize (2023 & 2024). Joinal's presentations, workshops, and publications primarily cover the following technical domains: Generative AI (GenAI) & LLMs: AI Agent architectures, RAG optimization (Vector Quantization, Semantic Caching), LLM scaling (Ray, Kubernetes), and ethical/responsible AI practices. MLOps & Cloud Native: Building scalable ML platforms using Kubernetes (KubeFlow, GKE), GitOps (ArgoCD, Flux), and managing high-throughput inference workloads. Cloud Architecture & Security: Implementing Zero-Trust architectures for AI, Kubernetes Security, and optimizing cloud services (GCP, Azure) for machine learning workloads. Relevance to Google Technologies: Google ADK, Vertex AI Agent Builder, Vertex AI Search, Dialogflow, Gemini API, Finetuning and Deploying Gemma, MLOps on VertexAI, GKE, Voice Agents using Gemini Live, ABAP SDK for Google Cloud, MediaPipe, WebGPU.

The industry is reaching a consensus: Agent = Model + Harness. While the LLM provides the reasoning, the harness provides everything else—the state, the tools, and the memory. But as model providers move toward closed, stateful APIs, a critical risk has emerged. If your agent’s memory is locked inside a proprietary harness, you aren't building a product; you are renting a relationship. 

This talk explores why Harness Engineering is the most critical discipline for AI practitioners in 2026. We will break down the anatomy of a harness and explain why memory isn't just a plugin—it is the harness. 

Read More →

Building AI agents is easy — building reliable AI agents is the hard part. This hands-on workshop is designed for data scientists, ML engineers, and AI practitioners who want to learn how to properly evaluate, test, debug, and deploy production-ready AI agents. 

Throughout the workshop, participants will work with a realistic customer-service AI agent using Google’s Agent Development Kit (ADK) to understand how to inspect agent reasoning, validate decision-making, detect failures, and automate regression testing workflows. Rather than evaluating only the final response, the workshop focuses on measuring the full sequence of actions, tool usage, and reasoning steps taken by an agent. 

By the end of the session, participants will build a complete automated evaluation pipeline — including verified test cases, bulk regression testing, debugging workflows, and deployment checks that prevent broken agent logic from reaching production. 

The workshop is heavily hands-on, with live coding exercises, realistic datasets, and end-to-end workflows that participants can directly apply to their own AI agents after the session.

Key Learning Outcomes 

  • Understand how to evaluate AI agents beyond final answers  
  • Build structured test cases from real agent interactions  
  • Run automated regression tests and debug failures  
  • Detect common production agent failure modes  
  • Set up deployment checks for reliable agent releases 
Read More →
View All Speakers

We use cookies essential for this site to function well. Please click to help us improve its usefulness with additional cookies. Learn about our use of cookies in our Privacy Policy & Cookies Policy.

Show details