Most of what I do falls into the six buckets below. But I don't sell a fixed menu. Every project starts with a conversation about what you're trying to do and whether AI, or a custom app, is even the right tool for it.
You're working with one person who has actually built this stuff, so you skip the layers of account managers and get straight, plain answers. I'm based in the New River Valley and work with clients wherever they are.
See how pricing worksIf you're not sure where to start, start here. I look at your operations and data, find where AI could actually earn its keep, and hand you a prioritized plan, including the parts not worth doing.
The core of the work: building models that predict, classify, or flag the things you care about, trained on your own data so the output reflects how your business actually behaves.
A model nobody can open isn't worth much. I build the whole application around it. Web and mobile software your customers or team actually use, with AI wired in where it genuinely helps rather than bolted on for show.
The boring, repetitive work, like sorting documents, copying data between systems, and answering the same email, is exactly what software is good at. I find those tasks and take them off your team's plate.
As a working academic, I help research teams add credible AI and synthetic-data methods to their proposals, and build the proof-of-concept models that make a grant panel believe you can deliver.
For creators and influencers: I dig into your engagement data to find what your audience actually responds to, so your content and growth decisions are based on evidence instead of a hunch.
Realistic artificial datasets that preserve the statistical properties of real data while ensuring privacy and compliance.
Specialized AI models fine-tuned for your domain, going beyond generic pre-trained models to deliver superior performance.
Strengthen proposals with AI methodology consultation, technical writing, and custom model development for NSF, NIH, and other agencies.
Strategic guidance on whether to build custom AI in-house or use existing platforms, with a clear analysis of costs and trade-offs.
Predictive modeling for patient outcomes, synthetic medical data, and workflow optimization with full HIPAA compliance.
Accreditation automation, student-success prediction, intelligent grading assistance, and administrative optimization.
Customer segmentation, inventory optimization, demand forecasting, and personalized marketing.
AI for creators to optimize engagement, analyze audiences, and build data-driven growth strategies.
Custom datasets that preserve privacy and statistical properties for testing, training, and research.
Cost-benefit analysis and roadmaps to decide between custom development and existing platforms.
We talk through what you're trying to do, look at the data you have, and decide whether this is even an AI problem in the first place.
I write up what's worth doing, in what order, with a rough sense of cost and payoff. You decide what's in.
I work out how it fits your existing systems (which model, where the data comes from, how it connects) before writing the real code.
I build it, test it, and get it running, then train your team so they can use and maintain it without me.
Once it's live we watch how it performs and tune it. A model that was right last year isn't automatically right today.
For specific AI initiatives with clear scope and deliverables.
For ongoing AI strategy and implementation support.
For strategic advice and knowledge transfer.
A general-purpose model is trained to be okay at everything. A fine-tuned one is trained on your data to be good at one thing. In practice that means:
It varies by scope, complexity, and duration. Typical ranges:
Those are starting points. I'll give you a real number once I understand what you actually need.
Tell me a bit about it and I'll let you know if it's worth pursuing, and roughly what it would take.