Devin BostTechnology advisoryLet’s talk

INDEPENDENT PERSPECTIVE. HANDS-ON DEPTH.

Technical judgment.
Built on delivery.

I help investors, founders, and enterprise leaders see the risks, understand the economics, and build technology that works in the real world.

Technical due diligence   /   Enterprise AI   /   Architecture

18 years

Engineering and architecture experience

10,000+

Users supported by an enterprise AI platform

$40M+

Annual revenue supported by a streaming platform

EXPERIENCE BUILT AT

IBMDataStaxOverstock.com (now Bed Bath & Beyond)Imagine Learning

01 / SELECTED WORK

The evidence behind
the advice.

From the investment decision to the production system. A few examples of where judgment changed the path forward.

INVESTOR DECISION SUPPORT01

A smaller scope.
A clearer investment.

Education technology startup · Technical feasibility

Assessed architecture, staffing, timelines, and cost, then presented the findings directly to an angel investor.

THE RESULT

The investor and CEO narrowed the product to a simplified assessment application.

The work behind the decision

The assessment connected the proposed product to the roles, team structure, engineering effort, and budget needed to deliver it.

Why it matters: A useful assessment can change the investment thesis before more resources are committed.

ARCHITECTURE & RELIABILITY02

Enterprise scale.
Operational confidence.

Overstock.com (now Bed Bath & Beyond) · Streaming platform

Jointly led platform evaluation and architected a messaging fabric serving hundreds of projects and billions of daily messages.

THE RESULT

Supported $40M+ in annual revenue. Reduced on-call incidents by approximately 99.8% through tracing and reliability improvements.

The work behind the outcome

Selected Apache Pulsar following an enterprise platform evaluation, co-authored architecture standards and the core Java framework, and trained teams on observability and adoption.

Why it matters: Platform selection and implementation have to hold up to the same scrutiny.

ENTERPRISE AI DELIVERY03

Beyond the demo.
Into the enterprise.

IBM · Joint IBM/NVIDIA customer

Architected and built an enterprise knowledge assistant, owning the core vector API and retrieval and agent workflows.

THE RESULT

A platform supporting 10,000+ users, 184,000+ documents, and 9.7 million retrieval chunks.

The work behind the platform

Supervised AI and Software engineers through technical direction, architecture and code reviews, implementation guidance, and mentoring.

Why it matters: AI readiness depends on the data, architecture, and team that make the application work.

The most valuable finding may be what you shouldn’t build. In my own AI startup, I identified more than $100K in unbudgeted costs required to validate the business model, costs that would recur with each new AI model.

02 / HOW I CAN HELP

Start with the decision
you need to make.

Focused advisory for moments when a technical assumption becomes a business consequence.

03 / THE APPROACH

Close to the business.
Deep in the details.

01

Begin with the business question.

What decision is on the table? Which constraints matter? I connect technical investigation to the outcome you need.

02

Follow the evidence.

Architecture, stakeholder interviews, prototypes, benchmark design, and operating costs reveal what a polished demonstration can miss.

03

Leave a path to execution.

Clear findings, explicit tradeoffs, and practical next steps—with the engineering depth to help carry them through.

04 / ABOUT DEVIN

An advisor who
still builds.

Devin Bost

Devin Bost · Technology advisor & software architect

I’m Devin Bost, a technology advisor and software architect with 18 years of experience connecting consequential decisions to working systems.

My work spans investor-facing feasibility assessments, enterprise data platforms, and AI delivery in high-security and regulated environments. I’m currently an AI Architect at IBM, following roles at DataStax, Overstock.com (now Bed Bath & Beyond), and Imagine Learning.

I’ve presented findings to investors and CEOs, interviewed more than 40 stakeholders for a single data-platform assessment, and supervised engineering teams through implementation. I’m comfortable moving between the business question and the code that answers it.

“When it comes to data and machine learning, Devin is exceptional.”

05 / IDEAS IN THE OPEN

Explore the technical depth.

Public writing and teaching on architectures that operate at scale.

LET’S GET TO THE REAL QUESTION

What decision
is in front of you?

Tell me what you’re evaluating, where the uncertainty is,
and what’s at stake.