Xander M. Lopez
Los Angeles, CA | xandermlopez@gmail.com | xandermlopez.com
linkedin.com/in/xander-lopez | github.com/XanderXML-Bit | zendren.net
Summary
Software engineer and program manager with hands-on experience building applications, defining requirements, and coordinating delivery with technical and business teams. AI Software Engineer at the U.S. Department of the Treasury, designing, building, and maintaining applications with senior executives, engineers, and program offices. Founder and lead of Zendren LLC, setting product priorities, defining requirements, and planning releases, as well as designing, building, and deploying its software.
Experience
AI Software Engineer, GS-14
U.S. Department of the Treasury, Departmental Offices | June 2026 – Present
Office of the CIO / Treasury Common Services Center
- Design, build, and maintain internal and external-facing software applications, taking each project from business requirements and architecture through implementation, deployment, and ongoing operation.
- Assess the needs of customers, senior executives, and program offices, map their manual workflows, and translate those workflows into prioritized software requirements.
- Develop full-stack applications, write automated tests, and conduct code and security reviews, using tools including OpenAI Codex, VS Code, Git, and GitHub.
- Configure infrastructure with Terraform, deploy containerized applications with Docker, and operate cloud services on AWS.
- Incorporate Federal Information Security Modernization Act (FISMA) and Authority to Operate (ATO) requirements, National Institute of Standards and Technology (NIST) SP 800-53 controls, and Zero Trust principles into application planning and review.
- Coordinate technical work on small teams and larger cross-functional efforts, and evaluate architecture and implementation options against user needs in reviews with engineers.
- Demonstrate working applications to stakeholders, revise them based on feedback, and document technical decisions and operating requirements.
Founder & Technical Lead
Zendren LLC | September 2025 – Present
- Plan scope, releases, and dependencies for Zendren's trading-research software, AI-agent platform, and web applications.
- Design application architecture, APIs, and data flows that connect language and vision models to application logic and storage, with controls that let users review and correct model output.
- Write Python and TypeScript, review code produced by AI coding agents, and maintain version-controlled repositories and technical documentation.
- Evaluate model providers, integration approaches, and development tools for capability, cost, reliability, and security, using prototypes and failure-case tests before committing to a technical approach.
- Define what each AI agent may read, change, and execute. Enforce those limits with per-workflow permissions, scoped credentials, session isolation, and human-approval gates, and keep audit logs.
- Harden services that handle untrusted content and external requests with prompt-injection defenses, server-side request forgery (SSRF) protection, per-IP rate limiting, timing-safe credential checks, and CSP and HSTS headers.
- Build parameterized Vitest and pytest suites for document parsing, financial calculations, and API failures, using synthetic inputs to test edge cases and security defenses.
Recreation Assistant (previously Recreation Aide)
City of Los Angeles, Department of Recreation and Parks | May 2022 – June 2026 (on call since June 2026)
- Coordinated planning, staffing, logistics, and day-of operations for seasonal events serving 500 to 800 attendees.
- Troubleshot AV equipment, registration systems, and staff devices, and trained coworkers on digital tools. Created signage and flyers, and supported registration and communication for programs serving hundreds of families.
Founder & Manager
Imperion Property Management LLC | April 2025 – September 2025
- Founded and operated a short-term rental management business, setting up the company website, business accounts, vendor workflows, and administrative processes.
- Recruited and managed two contract staff, coordinated schedules and service delivery, built a client pipeline through direct outreach, negotiated service contracts, and resolved client issues.
Independent Developer, AI and Software Projects
Self-directed | August 2020 – March 2025
- Built a Python chatbot that used retrieval-augmented generation with open-source language models to answer questions about UC San Diego Associated Students policies from official documents, and deployed it live for staff and students.
- Developed GPT-based financial-analysis tools, browser-automation agents, a self-hosted Stable Diffusion server, and early agent prototypes that led to Zendren's agent systems.
- Designed and administered a survey of 349 high school students on AI familiarity and career concerns, analyzed the quantitative and qualitative responses, and presented recommendations for AI education.
Projects
More work at xandermlopez.com.
Imperion. Multi-agent operations system. Zendren's multi-agent platform that runs scheduled operations and long-running engineering work across Hermes Agent, OpenAI Codex, and Claude Code. Layered memory combines an Obsidian knowledge base with hybrid keyword and vector search, a self-hosted conversational memory service that derives facts about users and projects, and a self-improving skill library that agents create, test, and refine from their own work. Connectors give agents governed access to email, calendars, accounts, CLIs, and developer tools, and applications are built agent-first so AI agents can operate and test them at scale. Model routing with automatic fallback and MCP tool access run under scoped credentials, human-approval gates, audit logs, health monitoring, and cross-provider code review.
AlgoTrade. AI trading research and execution platform. A large Python codebase that gives AI agents the tools, data, and workflow to make and manage trading decisions. An MCP tool server and custom scanners supply market data, news and headline capture, technical levels, multi-model consensus, and broker access. A strategy pipeline takes ideas from hypothesis to pre-registered, cost-adjusted backtests against market benchmarks to paper-trading leagues before anything goes live. In production, an agent trader runs a live brokerage account under hard risk controls, while a separate director agent reviews every trade, scores results against benchmarks, and evolves the prompts, tools, and strategies.
Technical Skills
Languages: Python, TypeScript, JavaScript, SQL, C/C++, C#, Rust, Lua, HTML/CSS
Web & Infrastructure: React, Node.js, REST APIs, Capacitor, Firebase, AWS, Terraform, Docker, Linux
AI: Claude Code, OpenAI Codex, Gemini CLI, Hermes Agent, open-weight models, LLM integration and evaluation, OCR, vision-language models, retrieval-augmented generation, vector databases, agent orchestration, model routing, MCP, prompt engineering
Engineering: Git, CI/CD, Vitest, pytest, API testing, code and security review, FISMA/ATO and NIST SP 800-53, technical documentation
Design: UI/UX, Photoshop, Canva, video editing
Education
University of California, Los Angeles | September 2026 – Present
Cognitive Science
Los Angeles Pierce College | September 2024 – June 2026
Computer science and general-education coursework. LACCD cumulative GPA 3.71. Dean's Honor List, President's Honors, Pierce College Honors Program, and Alpha Gamma Sigma honor society.
University of Southern California | Summer 2024
AI & Semiconductor Chips Summer Program, certificate