reducibl

a product team of one. live, on demand applied ai strategy and engineering

from idea to production in weeks

most AI projects don’t fail because of the models. they fail because nobody owns identity, permissions, or the path to production.

recent builds

previously aws, dell · harvard, uchicago · github · linkedin · email

ai sprint sessions

i build AI products with founders and teams — live, on-camera, collaboratively, in focused sessions.

from zero to working production system in weeks, not months.

you walk away with a real system your team can use immediately. not a prototype, not slides. ready to deploy or iterate on.

best fit for founders or small operating teams – including those in enterprise and pe-backed environments – serious about shipping ai.

i work equally well with technical and non-technical clients.

book an intro call — 45 minutes, no cost, we’ll build together — live

fixed-price packages

i also offer fixed-price packages for scoped deliverables starting at $10k

assess → build → harden

assess

genai initiative prioritization
stop guessing which AI projects to fund. source, score, and prioritize as a portfolio

ai architecture audit
a focused review of your AI stack with a prioritized action plan

build vs buy analysis
a structured analysis of whether to build or buy AI capabilities — with cost models and a clear recommendation

build

idea-to-pilot sprint
one week to turn your AI idea into a working pilot — validated, functional, ready to demo
see how this worked for apprentice →

mcp server build sprint
zero to production mcp server in one week. deployed, documented, yours
see how this worked for apprentice →

pilot-to-production sprint
two weeks of half day sprints to ship your stalled AI pilot — or kill it with a clear post-mortem

harden

ai governance layer setup
identity, permissions, and audit trails for your AI systems — powered by gatewaystack
see how this worked for inner →

not sure where to start?

you're spinning up a genai program and need to decide what to prioritize → genai initiative prioritization
you want to design and architect an internal ai platform or product → ai architecture audit
you're deciding whether to build or buy → build vs buy analysis
you have an AI idea that needs to become real → idea-to-pilot sprint
you need AI tools talking to real systems → mcp server build sprint
you have a pilot that's stalled → pilot-to-production sprint
you're worried about identity, permissions, audit → ai governance layer setup

not ready for a package?

start with a single sprint session. we'll pick one problem, build live, and you'll see how i work — no commitment beyond the session. if it leads somewhere, great. if not, you walk away with working code. → book an intro call

still not sure?

book a slot → weekly office hours

what i’ve shipped

inner — emotional memory for LLMs, available as a chatgpt app and on iOS

apprentice — AI-powered art study across 37,000+ masterworks

both run on gatewaystack, the trust and governance layer i built for user-scoped AI access. it handles identity, permissions, and auditability.

writing

apps are the flywheel, data is the asset — building a portfolio of AI apps? bet on the data

the three-party identity problem in mcp servers — the architectural challenge every agent system hits

is claude code secure? — secrets, prompt injection, and the real weak link

all writing →

build logs

notes from building AI-native systems in public.

latest logs →