About Me & Background
With years of experience delivering AI and ML solutions that scale—from scrappy startups to global enterprises, I can help you build, deploy, and scale your AI/ML solutions. I also specialize in helping engineering teams become AI power-users, implementing tools, workflows, and agents that measurably accelerate delivery.
Expertise
- AI/ML System Architecture
- AI Developer Productivity & MLOps Acceleration
- Custom AI Agents for Engineering Workflows
- Deep Learning & NLP
- Data Science & Analytics
- Technical Leadership & Mentoring
- AI Product Strategy
- Workshops & Public Speaking
Featured Projects
Minevra (jobs.minevra.co)
Your career, centralized. Seamlessly organize, track, and optimize your job applications—all in one place.
View ProjectWhimsy Quiz (whimsyquiz.minevra.co)
A quirky, AI-powered quiz app serving up an endless stream of 'Which ____ are you?' results. No grown-ups allowed!
View ProjectAgent Systems
Built agent systems for clients, using multi-agent architectures to solve complex problems.
Cold Hard Cache
My personal blog. Proud winner of 'Least Visited Corner of the Internet' several years in a row. You're welcome to change that.
View ProjectApache Spark
Contributed meaningful user-facing changes to Apache Spark SQL 4.0+ (example #1, #2), along with authoring a new benchmarking suite and much of a new unit testing framework.
I'm passionate about helping businesses like yours succeed with artificial intelligence. Let's explore how my expertise can align with your goals.
Turbocharged Engineering Pilot — $2k, 12 Weeks, Real Results
Six short sessions to find and implement AI-powered productivity wins for your engineering team — with a money-back guarantee.
6 × 20-min sessions over ~12 weeks
AI Productivity Audit scorecard
5–10 Quick Wins with step-by-step "how to"
90-Day Roadmap for sustained gains
Optional micro-agent build-out
Offerings
Current Availability as of September 24, 2025: I'm currently booked for the next 4 months. New six-month engagements can begin in January 2026 (I can offer a warm up period as it approaches, as I am eager to start working with you). The Turbocharged Engineering Pilot also has limited availability—reach out to secure your spot.
Most consultants will promise you the world in 2 weeks. I won't. Real transformation and value creation takes time. I'm proposing 6 months (at a minimum) because that's how long it actually takes to understand your problems, figure out what would actually help, build and/or set up the solution properly, test it with real usage, iterate based on what we learn, train your team to maintain it, and actually see sustained results. If you want working systems that your team will choose to use on a daily basis, that are still up and running three years from now, let's talk.
RAG Excellence Program — 6-Month Engagement
$35k–$60k
Build a RAG system that actually works—reliably, on your data, and across your company.
Target Outcomes:
Dramatic reduction in support volume through intelligent self-service knowledge base
Accelerated cross-team collaboration via unified knowledge platform (Warning: Your team will become more valuable as their AI skills grow. Budget accordingly—they'll command higher salaries.)
Quick wins within weeks, followed by continuous delivery of high-impact results
Result: A company-grounded RAG platform backed by your data, genuinely usable across the organization.
Production RL Infrastructure — 6-Month Build
$70k–$90k
Ship RL that actually works in production.
Result: RL that actually works in production.
AI Integration Layer — Custom Engagement
Monthly retainer + billable hours
Make AI calls feel like any other function call in your codebase.
Target Outcome:
Engineers ship AI features 3x faster with reusable patterns
Stop treating AI as "special"—integrate it properly so your engineers can use LLM calls as naturally as they call any API. I'll build a production-ready abstraction layer directly into your codebase:
Type-safe wrappers with proper error handling, retries, and fallbacks
5–10 reference implementations across your stack so engineers learn by example
Testing patterns that make AI-dependent code easy to test and mock
Documentation that gets new engineers productive immediately
Ideal for: Teams who want AI deeply integrated into their development workflow, not bolted on as an afterthought.
FAQ
Why 6 months when other consultants promise results in 4 weeks?
Because those 4-week projects usually break in production or get abandoned. Real AI infrastructure needs time to integrate with your actual data, handle edge cases your team discovers in production, and train your people to maintain it. The 2-month exit points mean you're not locked in—but if we're seeing value, we keep going until it actually works.
What if we want to start smaller?
That's exactly what the Turbocharged Engineering Pilot is for—$2k, 12 weeks, real results. You get quick wins, a roadmap, and a clear sense of how we work together. If it clicks, we can scale up to one of the 6-month engagements.
12 Weeks for 6 sessions?
The 12-week schedule is intentionally flexible—we can front-load sessions if needed and still finish with time to spare. This relaxed pace gives you space to gather materials that will make our work more effective, and gives me time to deeply consider your challenges between sessions.
Do you work with teams that don't have production AI yet?
Yes, but you need to be ready to iterate. If you're expecting a vendor to deliver a turnkey solution without your team's involvement, we're not a fit. These engagements work best when your engineers are actively learning and collaborating throughout.
What happens after the 6 months?
Your team owns it. You'll have documentation, runbooks, monitoring, and engineers who know how to maintain and improve the system. I can stay on for ongoing support if needed, but the goal is always self-sufficiency.
How do you measure success?
We define metrics together at the start—things like response accuracy, adoption rates, time-to-production for AI features, or support ticket reduction. Every 2-month checkpoint includes a review of these metrics. If they're not moving in the right direction, we adjust or part ways. But here's the reality: metrics are a guiding light, not the finish line. A truly successful project ends with both of us looking back and telling the same story—that it worked, that it mattered, and that your team is better off for it.
Let's Chat
Want to start the Turbocharged Engineering Pilot? Fill out the survey and book your spot here.