engineer & ai researcher
eric
brown.
you know the brands.
meet the engineer.
i build the go services behind real products and go deeper into the architectures behind ai.
the person behind the code 01 / a few things i’ve worked on
familiar on the outside.
my work is inside.
eight chapters. eight objects.
hover to open. select for the story.
from backend apis to ai architecture.
the services behind
the experience.
the architecture behind
the intelligence.
02 / independent ai research
open the model.
follow the evidence.
what’s actually in a model? what did it learn? what can we prove?
my independent research practice
optrenium.ai
my research into neural architectures,
learning systems, and model internals.
learning from scratch.
evidence you can inspect.
model audit & evidence
probity
byte-level gguf auditing that turns model internals into inspectable evidence. content-addressed claims, tensor and attention-head analysis, expert structure, and traceable inspector gadget stages.
learning engine
acuity
a proprietary learning engine that fits each uploaded dataset from scratch. its live demo reports held-out quality, a shuffled-target control, and calibrated regression intervals.
forge
evidence-bound planning for model reconstruction. fits plans to file or hardware constraints, accounts for runtime reserves, and keeps static reconstruction evidence separate from behavioral evaluation.
04snapper
a portable, git-aware tool for capturing source projects as compact text snapshots for llm prompts, with support for reconstructing files from snapshots.
05telos benchmarks
public benchmark results and evaluation data for telos, an experimental ai architecture.
03 / the person behind the code
engineer.
researcher.
builder.
working with teams everywhere.
my father built the computer.
i wanted to understand it.
in 1981, my father built a zenith heathkit h89. by seven, i was teaching myself basic. that early fascination became a lifelong drive to understand how things work and to build them myself.
from c and c++ to go in late 2013, i’ve always been drawn to the machinery beneath the abstractions. today, that means neural architectures, learning systems, and the mathematics behind intelligence.
i bring that same curiosity to engineering teams: clear thinking, honest evidence, practical mentorship, and systems that hold up after they ship.
a heathkit h89 at home.
basic. age seven.
c, c++, and systems.
simplicity. concurrency.
first-principles ai research.
the résumé / every chapter
the full record.
roles, responsibilities, and the technologies i worked with.
support enterprise consulting engagements across api integration, system design, and ai development. updated api systems for capital one’s discover merger, adding support for discover card metrics and related integration requirements. validated merger-related changes through a coordinated, multi-week manual testing effort covering end-to-end api workflows. currently work with amazon on a proprietary calculation tool with integrated ai pipelines. provide technical direction through architecture discussions, implementation, code review, and collaboration with client engineering teams.
“he is eager to learn and faces any challenge head-on. any team would be lucky to have him on their side.”mike deremusformer teammate · linkedin recommendation, june 2024
04 / notes from the workbench
a place to show
the working.
a new journal on model architecture, go, and what happens when an experiment meets reality.
open the journalhow things
work_
open to the right challenge
have a hard
problem?
let’s build.
ai architecture, backend apis, or a team that needs an experienced technical lead. tell me what you’re working on.
let’s start with your idea
start a conversationopens your email app with your brief. prefer linkedin? connect here ↗