Professional history / Résumé
Tianna McCoy.
Senior full-stack engineer building agentic development workflows and researching local AI inference infrastructure, with a background in software delivery, mentorship, industrial operations, and military logistics.
Usually in the U.S. Pacific Northwest. Remote, with in-person visits arranged in advance. Authorized to work in the U.S.
Independent product work
I built ART and Day Shift / tnsds.tech as the sole contributor to both projects, connecting product design, application architecture, and implementation. Between HOMEE and Atlas, I focused on self-directed learning, independent software development, and community mutual aid. My current focus is TNSDS Stations: researching practical local inference infrastructure for small and medium-sized businesses.
Day Shift · Agentic developer tooling
- Designed a human-agent contract: templated Markdown specifications with lightweight metadata let agents discover relevant context and continue work across sessions without relying on chat history.
- Built agent-facing CLI commands for discovery, readiness checks, review, and artifact updates, returning JSON next-step guidance based on workflow state.
- Defined human approval boundaries: agents are instructed to stop and ask before expanding scope, and accepting completed work remains a separate human decision.
- Used Day Shift in its own development: as of October 2026, planning records document 119 completed phases covering 1,162 tasks with implementation summaries. Progressed from supervising tasks to milestones and full phases, followed by my code review and user testing.
- Used Codex and GitHub Copilot with Day Shift. Kept shared workflow context in local project artifacts, with prompts and selected context sent to configured remote inference providers.
Career progression
Engineering, mentorship, and operations.
Local Inference Infrastructure Research
- Researching how to make local AI inference infrastructure viable for small and medium-sized businesses, with older laptops repurposed as local servers.
- Exploring Ubuntu Server deployment, model quantization, and shared inference for multiple users, with attention to hardware constraints and practical operation.
Infrastructure Automation Engineer
- Used Day Shift to build the foundation and initial monitoring workflows for a network monitoring system (NMS). The application remained in development when my role ended.
- Deployed the application to a virtual machine with development and staging environments.
- Implemented ping-based device health checks and graphs of health-check history.
Product Development & Community Mutual Aid
- Deepened software engineering knowledge through self-directed learning and hands-on development of ART and Day Shift as the sole contributor to both projects.
- Connected product design, application architecture, and implementation through independent project work.
- Contributed to community mutual aid alongside continued technical learning and development.
Software Engineer → Software Engineer III
Software Engineer III
January 2023 – September 2024
- Owned frontend and backend platform features and contributed to architecture, reliability, performance, and maintainability.
- Worked with product and operations stakeholders; mentored engineers through reviews, standards, and documentation.
Read the engineering exampleSoftware Engineer
October 2021 – January 2023
- Coordinated development across five cross-functional teams for complex product launches.
- Proposed catch-all handling for unrecognized insurance-data fields and partnered with my team lead on detection and recovery workflows. Related downstream processing-error reports stopped following deployment.
Technical Career Advisor / Lead Remote Software Engineering Tutor
- Coached and mentored struggling learners through guided questioning and tracing data through code, helping them connect concepts and work toward finding answers independently.
- Supported HTML, CSS, JavaScript, React, Node.js, MongoDB, and Express learning, with technical mentorship across 25+ group sessions.
- Authored standard operating procedures for tutoring workflows and structured problem-solving guidance.
Journeyman / Supervising Millwright
- Led multidisciplinary teams in industrial environments, coordinating logistics, safety, and precision maintenance.
- Trained apprentices and maintained operational standards under time-critical conditions.
Motor Vehicle Operator / Infantry
- Managed transportation logistics and equipment readiness in environments requiring coordination, discipline, and reliability.
HOMEE / Engineering in practice
Making insurance-data changes visible and recoverable.
The problem
Our team was reporting missing information needed to process insurance claims. We retained the raw insurance-company payloads, but could not confidently account for all of that information in the application’s structured data. I questioned whether we were receiving information that was not making it through ingestion; working with my team lead, we confirmed that was happening.
My contribution and collaboration
I proposed a catch-all field for data our existing schema did not recognize. My team lead and I worked together on the implementation and recovery process, drawing on their senior experience. Our immediate priority was making the information needed for claims available; investigating changes and cleaning up affected data became ongoing responsibilities.
Detection and recovery
A non-null catch-all field signaled that something needed investigation. A scheduled job notified my lead through Slack, a field in the UI let non-backend engineers surface the issue, and raw logs and database records supported manual review. We could then account for new or deprecated fields, update future ingestion, and write cleanup scripts to bring retained data into the updated structure.
The observed result
Following deployment, we stopped receiving the downstream processing-error reports associated with this issue. We actively monitored for roughly two weeks and continued passive monitoring afterward.
Kenzie / Mentoring in practice
Helping learners connect concepts and find their own answers.
Coaching as my core responsibility
At Kenzie, my coaching and mentoring work focused on struggling learners. I used guided questions to help them identify answers themselves. I called my broader approach to researching and finding answers ‘GoogleFu’: the goal was to help learners build a problem-solving process they could use beyond the current assignment.
The learning challenge
One learner was roughly two to three weeks behind their peers. As I recall, they were working on the JavaScript Tower of Hanoi project and struggling to connect concepts they had recently learned into a working solution.
Guided reasoning through the code
I focused on how data moved through variables and functions: what each function received, how it transformed the data, and what it returned. As the learner talked aloud about what they saw, understood, and did not understand, I could ask targeted questions to nudge their reasoning forward. I later recognized the connection to rubber-duck debugging, and I was nicknamed ‘RubberDuck.’
The observed result
Over a few sessions lasting roughly one to two hours each, the learner’s understanding improved enough for them to catch up with their peers. Explaining their thinking aloud made their understanding and uncertainty visible, so I could adapt my coaching to where they were getting stuck.
ART / Product judgment in practice
Letting community needs guide the technical work.
Working at the speed of trust
I recognized that I was a stranger to the communities I wanted to support. Trust needed to develop through useful, observable contributions. I designed ART’s foundation so communities could retain control and operate independently. As trust grew, I wanted to understand what functionality would help, without asking communities to give me more control over their data. Many communities cannot afford dedicated technical support, and I wanted to make that support accessible.
A practical information gap
Donated goods were spread across multiple locations, with only rough estimates of what each location held and how much was available. This made it difficult to coordinate getting supplies to the communities that needed them.
Building the useful feature
I built functionality into ART to document what goods were available and where they were warehoused. Once we knew what we had and where it was, coordinating where those goods needed to go became easier. The community work identified a need, and that need informed what I built into ART.
Contributing where the need was
The software supported distribution by making supplies easier to locate. Getting goods into affected communities also required people doing the physical work. When that became the immediate need, I stepped away from the technical work and joined distribution myself. My contribution followed the community’s needs, from providing information to helping move goods.
Education & training
Full-stack foundations.
July 2020 – July 2021
Kenzie Academy
Full-stack software development · Project-based training
Completed a 12-month frontend and backend curriculum. Led the CloseSpace capstone as Product Owner, building a MERN application with Mapbox integration, authentication, and REST APIs.
Certification & community
Continuing practice.
Graph Developer – Associate
Apollo GraphQL
Issued February 2022Digital Corps volunteer
Out in Tech
Web services for LGBTQ+ organizationsSoftware engineering volunteer
Code for America
Civic technology and public service
This page includes my professional history and subsequent updates. Use “Print / save as PDF” for the current résumé. Certification and volunteer details are listed on my LinkedIn profile (opens in a new tab) ↗.
Explore the fit
Connect the background to the work.
My hiring overview covers current interests and availability. ART and Day Shift show my independent product work; working expectations explain the collaboration I’m looking for.