Software Engineer & Applied AI Consultant · Owner, NextEleven

Mansfield, Texas / Dallas–Fort Worth · Available for remote consulting and software projects

NextEleven · Mothership · Discuss a project · Apple App Store · Google Play · YouTube · X · LinkedIn

I help businesses turn an unclear requirement, a broken workflow, or an unfinished application into a verified working system. My work spans backend services, business software, websites, Apple and Android applications, and applied AI.

You work directly with me: understanding the problem, choosing the approach, building or repairing the software, checking its behavior, and documenting how to run it.

• AI consulting and implementation: assess where AI can help a workflow, define the data and integration requirements, and build document search, source-grounded assistants, tool integrations, or workflows with human review. Establish evaluation cases before treating a useful demonstration as a dependable system.
• Custom software and integrations: build APIs, internal tools, inventory and order workflows, operator dashboards, customer portals, and business websites. Connect existing systems around a defined operational need.
• Software investigation and repair: reproduce failures, inspect logs and state, isolate causes, repair incomplete behavior, and improve tests, CI, deployment checks, and handoff documentation.
• Apple and Android application development: build and improve iPhone, iPad, Mac and Android applications, including platform integration, testing, store review remediation and release work.

An engagement can start with a focused assessment, a repair, or a defined build. We agree on the scope, acceptance criteria, and next decision before expanding the work.

Delivered a production web presence with English and Spanish journeys, separate buyer and seller paths, responsive interfaces, contact forms, and a bounded information assistant. The assistant uses approved business information and hands uncertain questions to the team. Inventory and ecommerce remain separate future scope.

Built a Python CLI, FastAPI service, and local interface for ingesting and searching files. The implementation includes keyword/vector retrieval, reciprocal-rank fusion, reranking, citations, REST/SSE access, and retrieval evaluation. Storage and retrieval run locally; external generation is an explicit configuration choice. MIT licensed.

Built a Rust admission gate with embedded, HTTP, and stdio MCP integration paths, operator revocation, and structured evidence. Published native releases for macOS, Linux, and Windows. Documented consumer tests exercise permitted and denied actions, outages, and recovery. Enforcement depends on the application's trusted dispatcher checking before execution. Apache-2.0 licensed.

Built a provider-neutral runtime with explicit model routing, repository tools, persistent sessions, bounded workers, terminal and local web interfaces, and MCP support. The acceptance record links cross-platform CI and distinguishes synthetic checks from live provider behavior. MIT licensed; current development focuses on dependable personal use.

Built a TypeScript/Next.js application, agent API, CLI, and local MCP adapter around pinned source, provenance, license notices, and reviewable component briefs. Controlled evaluations include fresh consumer integrations and recorded failures. A working development prototype; independent usefulness and demand remain unproven.

Apple App Store · Google Play

As checked on 2 November 2026: 12 current US App Store listings and 12 public US Google Play listings, including Dinnlume, Pairtrace and Circuit Breaker. Counts describe store publication, not downloads, adoption or revenue. Availability varies by region and device.

• Two completed confidential client systems: a multi-location dealership workflow system and a multi-component applied-AI system. Work included requirements, architecture, implementation, testing, demonstrations, and technical specifications. I can discuss architecture and delivery decisions within the confidentiality boundary.
• Witness: a Rust provenance prototype with signed records, SQLite storage, ingestion, and REST/GraphQL interfaces.
• EvidenceLens: experimental source verification and semantic-review evaluation, including preserved source versions, exact-span checks, reproducible scoring, and human-review protocols.

Project experience includes:

• Languages: Python, Rust, TypeScript/JavaScript, Swift, Kotlin, SQL, and Bash.
• Applications and interfaces: FastAPI, React, Next.js, SwiftUI, UIKit, Jetpack Compose, and React Native.
• Data and delivery: PostgreSQL/pgvector, SQLite, Redis, ChromaDB, Docker, GitHub Actions, automated testing, release checks, and runbooks.
• Applied AI: LLM APIs, tool calling, MCP, provider/model routing, document ingestion, hybrid retrieval, reranking, citations, evaluation, and human review.

I use AI tools throughout investigation, implementation, and review. I remain responsible for the architecture, the decisions, the verification, and the delivered result. I check the actual system and make limitations visible.

Before making software my primary professional direction in 2025, I spent 27 years in a client-facing profession, including eight years owning and operating a business. That experience informs how I discover requirements, set expectations, communicate difficult issues, and take responsibility for delivery.

Tell me what needs to work better, what you have today, and what a successful result would look like. You do not need a finished specification to begin.

hello@nexteleven.dev · NextEleven · Mansfield, Texas / Dallas–Fort Worth