I'm focused on making AI agents reliable. I co‑founded Zalor, where we automatically test agents before deployment. At ServiceNow, I build observability for AI agents. I define problems, set product direction, and ship end‑to‑end.
Zalor
Sep 2025 – Present
I founded Zalor after identifying a gap in how teams build and ship AI agents. Through 45+ customer conversations, I found most teams relied on manual testing of prompts, tools, and workflows and only discovered failures after deployment.
To address this, I built a testing platform for AI agents, including scenario generation, output evaluation, and regression detection. I've signed multiple paying customers and am working with teams deploying AI agents.
ServiceNow
Aug 2024 – Present
At ServiceNow, I work on AI Control Tower, where I lead research, product strategy, and development of integrations for multi‑agent frameworks. I partner with product and engineering teams to define the approach, build demos, and drive these integrations from concept to production.
I built connectors for frameworks like CrewAI and LangGraph, enabling teams to monitor and control their agents directly from the platform. I also developed integrations across AWS AgentCore, Azure AI Foundry, and GCP Vertex AI, allowing users to observe agent traces from a single, centralized interface.
I built the new AI‑native UI for AI Control Tower, redesigning the platform experience from scratch and presenting a product demo to 500+ employees ahead of release.
ServiceNow
May 2021 – Aug 2023
I interned at ServiceNow for three years across software engineering and quality engineering roles. During this time, I worked on the Mobile App Builder and Field Service Management platforms, building product features, improving test reliability, and resolving core issues.
I created product demos presented at ServiceNow's internal conference (UTG Connect) to over 2,000 employees, gaining early exposure to how products are communicated.
In my final internship, I won 1st place out of 500+ participants in ServiceNow's Engineering Hackathon by building an intelligent code editor for running scripts in the platform. After winning, we shipped the feature to production, where it is now used by thousands of customers.
AI Agent Testing Platform
AI agents are responsible for real actions like creating records, triggering workflows, and interacting with production systems, yet most teams still rely on manual testing and only discover failures after shipping. To solve this, I built Zalor, a platform that uses AI agents to test other AI agents. It generates high‑coverage scenarios, evaluates output correctness and tool usage, and catches regressions before they reach production.
Most teams rely on manual testing for AI agents and only discover failures after shipping. Zalor uses AI agents to test other AI agents, catching regressions before they reach production.
MCP Server Generation & Hosting
Building MCP servers is painful and time‑consuming, often taking over a month to set up and requiring ongoing maintenance to handle issues like context limits and redundant tool calls, with no reliable feedback loop after deployment. To address this, I built a platform that generates and hosts MCP servers from OpenAPI specs, enabling businesses to connect their APIs to AI assistants like ChatGPT and Claude. I partnered with Weav.ai and conducted 20+ customer conversations with companies like Monday.com, Outreach, and Kubera to validate demand.
Building MCP servers is painful and time-consuming. I built a platform that generates and hosts MCP servers from OpenAPI specs, enabling businesses to connect their APIs to AI assistants like ChatGPT and Claude.
ServiceNow Hackathon Project
The background scripts code editor is a core tool used by developers daily on the ServiceNow platform, yet it hadn’t been meaningfully updated in over 18 years and remained a basic text box without modern IDE features. To address this, we built Scribe, an intelligent code editor with real‑time error detection, automatic formatting, code caching, and auto‑completion. Scribe ranked 1st place in ServiceNow’s company‑wide engineering hackathon and was later shipped to production, where it is now used by thousands of developers.
The background scripts code editor is a core tool used daily on the ServiceNow platform, yet it hadn’t been updated in over 18 years. We built Scribe, an intelligent code editor with real‑time error detection, formatting, caching, and auto‑completion. Ranked 1st in ServiceNow’s engineering hackathon and shipped to production.
ServiceNow Hackathon Project
Developers are often tied to their workstations to track project progress and review changes. Fluid is a mobile application that provides real‑time access to project data, enabling developers to monitor user stories, track defects, and review pull requests directly from their mobile devices.
Fluid is a mobile app that keeps developers connected by enabling real-time tracking of stories, defects, and pull requests on the go.
ServiceNow Hackathon Project
Keeping track of task status across tools can be tedious, especially when juggling multiple workflows. Stub is a Microsoft Teams bot that enables users to create, assign, and track tasks directly within Teams, reducing context switching and improving productivity.
Stub is a Microsoft Teams bot that streamlines task creation, assignment, and tracking directly within team chats.
Founder of Nonprofit Organization
Access to quality education is often limited in underfunded school districts. I founded MotivateKids, a nonprofit focused on making education more engaging and accessible for underserved students. I built and scaled a team of 35 tutors delivering free tutoring, workshops, and mentorship programs to 1,200+ students across multiple schools in the Moreland School District.
MotivateKids is a nonprofit I founded to make education engaging and accessible for underprivileged students through interactive teaching and mentorship.
