TECHNOLOGY EXECUTIVE · CLOUD & PLATFORM ENGINEERING

I build cloud platforms that scale — and the teams that run them.

CTO and Director of Engineering across IoT, enterprise SaaS, and e-commerce — now integrating AI and ML into production platforms. Currently Director of Software Engineering & Technology at OmniMetrix.

View selected work Download resume (PDF)

Buford, GA · Open to remote · linkedin.com/in/matthewfirth

Matthew Firth
40K
IoT devices monitored
50%
cloud spend reduction
100+
Fortune 500 clients supported
150
person tech org led

01 — SELECTED WORK

Local-First Memory Layer for AI Assistants

MODELBRAIN · SOLO FOUNDER PROJECT · 2026–PRESENT
LOCAL VAULT AES-256 TAG + INDEX RECALL TEST MCP MEMORY SERVER ANY AI ASSISTANT FREE LOCAL · $.05/GB

PROBLEM

AI assistants forget everything between sessions, or remember only inside one vendor's cloud, priced by the gigabyte, with no clear guarantee about what "forgetting" a memory actually means.

APPROACH

A memory layer built local-first: an encrypted vault on the user's own device, a plain tagging scheme for what gets kept or deleted, and recall tested directly rather than assumed. Benchmarked against MCP-based memory servers and platform-native options. Free to run entirely locally, no account required; only optional cloud sync and backup are billed, at a flat 5 cents per gigabyte.

OUTCOME

A working, vendor-agnostic memory layer for AI assistants, documented across a dozen articles on pricing, encryption, and recall design, at modelbrain.net.

local-first storage · AES-256 vault encryption · MCP memory server · free local, paid cloud sync · recall testing

AI-Assisted Matchmaking Platform

NEXSPARK · SOLO FOUNDER PROJECT · 2025–PRESENT
QUIZ → PROFILE VERIFIED MEMBERS AI SHORTLIST HUMAN REVIEW ANON MATCH REPORT MUTUAL YES → INTRO AI GROUPING CIRCLE of 4–6 FIRST MEETUP planned NEXSPARK.COM · MATCHMAKING NEXSPARK.ORG · FRIENDS

PROBLEM

Dating apps optimize for engagement, not outcomes — endless swiping, unverified profiles, and incentives misaligned with users actually meeting someone.

APPROACH

A deliberate solo experiment in AI-assisted building, alongside the OmniMetrix role: a values-based quiz producing structured profiles, an AI pipeline that shortlists matches across identity-verified members in both directions, human review of every pairing, and anonymized match reports — identities revealed only on a mutual yes.

OUTCOME

Two live products from one matching engine — matchmaking at nexspark.com and group friend-matching at nexspark.org — designed, built, and shipped end to end by one person: AI pipeline, identity and age verification, messaging, and the marketing/SEO sites.

AI matching pipeline · identity & age verification · anonymized messaging · structured profiles · SEO/content platform

Industrial IoT Data Platform

OMNIMETRIX · DIRECTOR OF SOFTWARE ENG · 2023–PRESENT
IoT DEVICES telemetry TCP RTU SERVICE BUS MICROSOFT SQL SERVER ML / AI FORECAST PORTAL K8s·Angular·.NET

PROBLEM

A fleet of 40,000 industrial monitoring devices producing telemetry with no unified platform to store, analyze, or act on it — and cloud costs growing unchecked.

APPROACH

Designed a streaming ingestion pipeline — TCP RTU and Azure Service Bus feeding Microsoft SQL Server — with ML-driven forecasting and AI integrations (Azure OpenAI, Claude) on top, and a customer portal on Kubernetes. Instituted CI/CD with Cypress end-to-end testing and a NIST-aligned security program.

OUTCOME

14TB of analytics-ready telemetry. Cloud spend cut 50%. Audit-ready NIST compliance and automated release pipelines the team runs without ceremony.

Azure · SQL Server · Kubernetes · .NET Core · Angular · Service Bus · ML forecasting · Azure OpenAI / Claude

Legacy-to-Cloud Modernization

COMVIEW / CALERO MDSL · 2014–2023
LEGACY MONOLITH MICROSERVICES AZURE · K8s AZURE SQL REACT FRONT ENDS VENDOR APIs AT&T VERIZON SERVICENOW VMWARE

PROBLEM

A telecom expense management monolith serving demanding enterprise clients — aging, hard to scale, and expensive to extend across global billing regimes.

APPROACH

Incremental decomposition of the monolith into microservices — hosted in Azure, orchestrated with Kubernetes — with React front ends, deep vendor API integrations (AT&T, Verizon, ServiceNow, VMware), and SQL Server performance work with selective offload to Azure SQL.

OUTCOME

A platform trusted by 100+ Fortune 500 clients, with global billing compliance and internationalization built in — modernized without a big-bang rewrite.

Azure · Kubernetes · C#/.NET · SQL Server · Azure SQL · React · REST APIs · i18n

E-commerce Replatform & Automation

PETCARERX · 2006–2014
WEB FARM 01 WEB FARM 02 WEB FARM 03 COMMERCE PLATFORM WAREHOUSE AUTO PHARMACY AUTO 20,000+ pkg/day

PROBLEM

A growing pet-pharmacy retailer constrained by a slow storefront and manual fulfillment that couldn't keep pace with order volume.

APPROACH

Full storefront replatform tuned for <2s page loads on 800KB payloads, plus warehouse and pharmacy automation integrated directly with the commerce platform. PCI compliance throughout.

OUTCOME

Sales up 60% year over year after the redesign. 20,000+ packages a day through automation, 10k orders/day at peak — run by a 150-person technology organization.

.NET · SQL Server · web farm HA · warehouse automation · PCI

02 — LEADERSHIP

Built and scaled engineering and design organizations at three companies — from a handful of developers to a 150-person technology organization. Led Agile transformations that turned long, unpredictable release cycles into a steady sprint cadence — cross-functional teams where designers, engineers, and product plan and ship together, with CI/CD automation stripping the ceremony out of every release. Owned hiring and mentoring pipelines and ran vendor and outsourcing relationships across three continents.

Established security and audit programs to NIST, ISO 27001, SOC 2, PCI, and SOX standards — audit-ready, not audit-scrambling — and reported directly to executive leadership and boards on technology strategy, budgets, and risk.

GROOMING SPRINT PLANNING STANDUPS daily QA · SECURITY CI/CD PIPELINE RELEASE RETROSPECTIVE → feedback into grooming

WHAT COLLEAGUES SAY

"Working with Matthew was like having an architect and an engineer in the same room. He could quickly identify root causes, communicate solutions clearly, and guide implementation without losing sight of the business objectives."

BRENARD J. CONCARDO · Centric Corporation

03 — CAREER TIMELINE

2025—NOW
NexSpark FOUNDER
Founded human-driven, AI-assisted matchmaking (nexspark.com) and NexSpark Friends (nexspark.org) — AI matching pipeline, identity verification, full product build
2023—NOW
OmniMetrix
Director of Software Engineering & Technology — industrial IoT data platform
2014—2023
Comview / Calero MDSL
Director of Engineering — telecom expense SaaS for Fortune 500
2006—2014
PetCareRx
CTO — e-commerce replatform, warehouse & pharmacy automation, 150-person tech org
2001—2006
EuroRSCG (Havas)
Technology lead — digital agency builds for global brands
1998—2001
TPI
Senior engineer — enterprise systems
1995—1998
T-NET FOUNDER
Founded a commercial ISP in Memphis, TN
1993—1995
Wang (Thailand)
Systems engineer — first role in enterprise computing

Full detail in the resume PDF ↓

04 — CAPABILITIES

Cloud & Platform
Azure (deep) · AWS · Kubernetes · Docker · IaC · CI/CD
Data & AI
Kafka · Data Lake / Synapse / Databricks · ML forecasting · Azure OpenAI / Claude
Software
C#/.NET · SQL Server at scale · TypeScript · React / Angular / SolidJS
Security & Compliance
NIST · ISO 27001 · SOC 2 · PCI · SOX · audit readiness
Leadership
engineering & design teams · Agile · budgets · vendors · executive reporting
Infrastructure
datacenter → cloud migrations · networking · HA design

05 — FAQ

Is Matthew Firth open to new roles?
Yes. Open to VP Engineering, Director of Engineering, and CTO conversations, based in or remote from the Atlanta, Georgia area.

What does he do now?
Director of Software Engineering & Technology at OmniMetrix, running a 14TB industrial IoT data platform and a NIST-aligned security and audit program, with a team that cut cloud infrastructure spend by 50% without reducing capability.

What has he founded?
ModelBrain, a local-first AI memory layer for AI assistants at modelbrain.net, and NexSpark, a human-reviewed AI-assisted matchmaking platform at nexspark.com and nexspark.org.

What's his background?
25+ years in technology leadership across IoT, enterprise SaaS, and e-commerce, including CTO at PetCareRx and Director of Engineering at Comview / Calero MDSL, leading legacy-to-cloud modernization for 100+ Fortune 500 clients.