JOHNNY BRYANT
Eight-plus years across AI-enabled applications, APIs, backend services, full-stack systems and data platforms — building them, then debugging, reviewing and evaluating them for correctness, edge cases and real-world behaviour. Day to day that's |
I'm a senior software developer working remotely from Florida. Eight-plus years of it has been production engineering problems: APIs and backend services, full-stack systems, data pipelines, and lately AI-enabled applications at NerovixaAI. The interesting problems are rarely the pretty ones — data that has to stay correct, integrations with systems you don't control, and behaviour at the edges nobody wrote a ticket for. A good half of my week is spent reading code rather than writing it: review, root-cause analysis, and evaluating AI-generated code and technical responses for correctness, instruction adherence and failure modes.
Johnny Bryant · Senior Software Developer
Eight-plus years of production engineering: building the services and the product features, then reviewing, debugging and evaluating them until they behave the way the spec said they would.
Reviewing AI-generated code and technical responses for correctness, instruction adherence, clarity and failure modes.
Ask me about itRoot-cause analysis, defect investigation, edge-case and regression checks, and fixes that hold up in production.
Ask me about itAPI design and concurrent backend services in Python, Go, Java and TypeScript, plus the workflows that sit on top.
Ask me about itFull-stack products with AI-assisted processing behind them, from the interface down to the services.
Ask me about itETL/ELT in Python and SQL, data modeling, validation and reconciliation into PostgreSQL and AWS analytics.
Ask me about itSystem integration with the platforms you don't control, and the unglamorous work of keeping it all steady at scale.
Ask me about it
Python, Go and Java services behind AI-enabled products — turning product requirements into reliable production software and service workflows.
Checking generated code and technical responses for logical correctness, instruction adherence, failure modes and practical engineering quality.
TypeScript, Java and C# web applications, APIs and backend services supporting high-volume restaurant ordering and operational workflows.
Root-causing menu, order, customer-data and third-party API failures in Java and TypeScript services, then turning them into regression checks.
Full-stack services and API-connected features in Go and C++ for transaction-oriented hospitality workflows — ordering, payment and operations.
Python and SQL pipelines consolidating KYC, fraud and onboarding data from Alloy's API and third-party providers into PostgreSQL and AWS analytics.
You tell me what the system needs to do. I ask a lot of questions — probably more than you expect.
A short written plan: what I'll build, the trade-offs I can see, and roughly how long it takes.
You see it as it comes together, so if something isn't right we change it early.
Shipped, documented, and handed over properly. I'm still around when something breaks at 5pm on a Friday.
AI-enabled applications, APIs and backend services in Python, Go and Java. I review and debug application code and AI-assisted technical outputs for correctness, maintainability, edge cases and API-contract issues, build reproducible engineering scenarios to validate solutions, and apply AI/LLM evaluation practices — instruction adherence, failure modes, clarity — to technical workflows.
TypeScript, Java and C# web applications, APIs and backend services for high-volume restaurant ordering. Code review and structured debugging across every application layer, production failure investigation across menu, order, customer-data and third-party API workflows, and recurring issues converted into concrete regression checks.
Full-stack services and API-connected features in Go and C++ for transaction-oriented hospitality workflows spanning ordering, payment and restaurant operations — integration failures, inconsistent application state, data-flow defects and latency-sensitive behaviour across service boundaries.
Python and SQL pipelines consolidating identity-verification, application and decision data from Alloy's API and third-party providers into PostgreSQL and AWS. Modeled KYC, fraud and onboarding data for banks and fintech customers, with validation and reconciliation checks, operational dashboards, secure PII handling and audit-ready lineage.
Software engineering, algorithms, data structures, databases, web and full-stack development, and application design. ACM, Programming Club, Computer Science Club and a fair number of hackathons.
The fastest-moving part of my job: reviewing AI-generated code and technical responses, checking instruction adherence, and validating solutions against realistic engineering requirements.
Root-cause analysis, test-case design, edge-case and regression validation. Most of the value I've added has come from making existing systems steadier rather than building new ones.
14Aug
Half of restaurant tech is talking to systems you don't control. Notes on retries, idempotency and failing loudly enough.
Read it
02Aug
Instruction adherence, plausible-but-wrong solutions, and the failure modes that only show up once you run the thing.
Read it
21Jul
Root-cause analysis on transaction-oriented systems: turning a production incident into a reproducible case and a regression check.
Read itEmail is usually quickest, but the form works too. Whether it's a role, an evaluation or review engagement, or a system that's misbehaving, a rough outline is plenty to start with — I'll reply within a day or two, and I'll say honestly if it isn't something I should be taking on.