Speaker

Muzeeb Mohammad

Muzeeb Mohammad

Senior Manager of Software Engineering

Newark, Delaware, United States

Actions

Muzeeb Mohammad is a Senior Manager of Software Engineering at JPMorgan Chase, where he designs and leads large-scale, secure, and high-performance distributed microservices powering mission-critical financial platforms.

His work spans Zero-Trust API architectures, event-driven systems, and real-time data-streaming platforms (Kafka) supporting enterprise-scale credit-card and digital-banking services. He specializes in building systems that remain resilient, observable, and secure under extreme scale and regulatory constraints.

Beyond delivery, Muzeeb is a passionate engineering leader and mentor, focused on growing high-performing teams and developing the next generation of system architects. A Senior Member of IEEE and an international technology awards judge (including the Globee® Awards), he actively bridges academic research with real-world enterprise implementation.

Area of Expertise

  • Information & Communications Technology
  • Real Estate & Architecture

Topics

  • distributed systems
  • Microservices Architecture
  • FinTech Engineering Innovation
  • Cloud Native Design
  • Event Streaming (Kafka)
  • api security
  • AI Reliability Engineering
  • Resilience Patterns
  • Java
  • Spring Boot
  • Financial Systems Modernization
  • Spring
  • Spring Framework
  • Spring Cloud
  • Spring AI
  • Spring Batch
  • Spring Data JPA
  • Terraform
  • aws
  • CI CD
  • DevOps
  • Software Deveopment
  • software engineering
  • Distributed Software Systems

Designing Human-in-the-Loop AI Agents: Where Autonomy Must Stop

As AI agents become more capable, the biggest challenge is no longer intelligence—it is control. Fully autonomous agents can act fast, but without boundaries they introduce risk, unpredictability, and loss of trust.

In this session, we explore human-in-the-loop design patterns for AI agents, focusing on when and how autonomy should pause, escalate, or defer to humans. Using real-world inspired agent workflows, we demonstrate how agents can reason, plan, and propose actions—while humans retain authority over high-impact decisions.

The talk covers practical patterns for approval gates, confidence thresholds, action classification, and fallback behavior, showing how to build agents that are powerful yet safe to deploy. Rather than abstract theory, this session emphasizes concrete design choices and lessons learned from production-inspired systems.

Attendees will leave with a clear framework for deciding what agents should do, what they should recommend, and what must always require human judgment.

Building Agentic Java Microservices for Mission-Critical Systems

AI agents are rapidly moving from experimentation to production, but mission-critical Java systems demand far more than simple LLM integrations. Reliability, security, auditability, and controlled execution are non-negotiable.

In this session, we explore how to design agentic Java microservices that safely operate in production environments. Using a real-world inspired incident-triage use case, we demonstrate how Java applications can leverage AI agents to classify incidents, identify likely root causes, and generate remediation plans—while enforcing strict guardrails, policy controls, and human-in-the-loop approvals.

We will walk through an end-to-end architecture built with Spring Boot, Kafka, and modern Java AI frameworks, showing how AI agents can reason over logs, metrics, and traces without being granted unrestricted autonomy. The session emphasizes Zero-Trust principles, deterministic fallbacks, and observability-first design to ensure that AI-driven decisions remain explainable, auditable, and safe.

Attendees will leave with practical design patterns, architectural guidance, and lessons learned for building production-grade, agentic Java systems that enhance operational resilience rather than introduce new risk.

Muzeeb Mohammad

Senior Manager of Software Engineering

Newark, Delaware, United States

Actions

Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.

Jump to top