AlignTrue
AlignTrue is an open-source reference architecture for supervised AI systems that need durable receipts, replayable state, governed actions, and fenced side effects.
Learn moreReceiptsReplayabilityGovernance
Applied AI systems for operational teams.
Main Character builds retrieval, workflow, and review tools for teams that need software they can inspect, operate, and improve under real constraints.
Current work centers on source-aware retrieval, evaluation, wearable interfaces, and human review controls.
Operating model
Evidence, judgment, action, and records stay connected.
01
Evidence
Source material, policy, state, and prior decisions stay close to the task.
02
Judgment
People see the context, tradeoffs, and review points before action.
03
Follow-through
Actions leave durable records that teams can inspect, replay, and improve.
Software for teams that need accurate answers, usable controls, and inspectable records.
Useful AI products need more than a model call. They need source evidence, workflow state, human review, and interfaces that fit the operating environment.
We build around the process: documents, permissions, decisions, exceptions, handoffs, and the records teams need after action.
The aim is lower cognitive load and better execution: less searching, clearer next steps, and systems teams can inspect and improve.
Selected projects
Work across AI operations infrastructure, model governance, and frontline maintenance support. Different domains, same pattern: context, review, action, records.
These are selected public examples, not an exhaustive project history. Some related systems, client implementations, and reused patterns remain private because they are not appropriate for public detail.
GlassAlpha
GlassAlpha is an open-source toolkit for deterministic audit reports for tabular ML models, with fairness, explainability, calibration, robustness, and reproducibility metadata in one local workflow.
Learn moreML auditFairness testingReproducible reports
Maintenance Control
A private MVP build for approved-publication retrieval, voice query, and source-cited maintenance guidance on phone, web, and wearable surfaces.
Learn moreGrounded retrievalVoice queryControlled manuals
Other projects
Additional work across ecommerce AI, SaaS automation, developer and CMS tools, documentation software, and product leadership for technical teams. Not every engagement or implementation is publicly listed.
About the teamEcommerce AISaaS automationDeveloper tools
Capabilities
The reusable parts of applied AI work: retrieval, review, evaluation, interfaces, and handoff.
Source-aware retrieval
Workflow orchestration
Human review controls
Evaluation plans
Operator interfaces
Wearable workflows
Audit trails
Deployment handoff
How we work
Small teams, clear constraints, fast loops. Prototype narrowly, validate with users, and leave behind software that can be maintained.
Prototype fast
Start with a narrow workflow, build the first usable version quickly, and test it against source material, edge cases, and user constraints.
Build around the work
Design around the operator, the environment, the permissions, and the decision path instead of forcing people into generic software patterns.
Ship useful systems
Move from prototype to maintainable systems with clear interfaces, traceable outputs, evaluation, and human review where it matters.
Applied R&D and government work
Main Character takes on selective applied AI work when the problem is concrete: grounded retrieval, workflow automation, operator tools, and prototypes that need traceable outputs, review points, and a path to deployment.
The prototype is only part of the work. The workflow, users, assumptions, and constraints need to be understood with domain context.
Aviation work can draw on Corrie Mays, a former Marine Corps aviator and Naval Flight Officer (NFO). Other domains should bring the right operators, analysts, maintainers, or reviewers.
Grounded retrieval prototypes
Human approval gates
Workflow audit trails
Evaluation plans
Operator discovery
Deployment handoff
About the team
Main Character is founder-led by Gabe Mays, with project-specific subject-matter review for work that depends on operational judgment.
Before Main Character, Gabe led product teams at GoDaddy and built products across SaaS, automation, developer tools, and AI systems. Before tech, he served as a Marine Corps Captain with deployments to Iraq and Afghanistan.
Aviation-adjacent work can draw on Corrie Mays for credible subject-matter review. Other domains should bring reviewers who match the workflow and constraints.
Talk shop
Ready to build?
Tell us briefly what you're building, modernizing, or trying to make easier to operate.
Good fits
Retrieval, workflow software, evaluation, applied R&D, and operational tools where context matters.
What happens next
Gabe will review and follow up directly if it's a good fit.