Skip to content

SYSTEMS & PROCESS ENGINEERING · AI FOR MANUFACTURING

We Rebuild How Your Operation Runs — With AI as the Operating Layer

BlackArc is a systems and process firm. We find where AI changes the economics of your operation, re-engineer how the work gets done, then rebuild it — agentic workflows that act, models tuned on your own data, and systems that improve themselves. Agents are a component, not the product.

No migrations required. No platform to buy. Under 20 hours of your team's time.

Days → Minutes

Decision cycles

Multi-system coordination that took days resolves in minutes — exceptions handled before they compound.

10×

Skill & tool output

Proprietary tuning methods lift the quality and reliability of AI skills and tools by an order of magnitude.

Weeks

To production

Scoped, eval-validated systems running in your environment in weeks — not 18-month roadmaps.

Your data

Your models

Custom models tuned on your operation — your style, your decisions — running where your compliance team needs them.

01 // WHAT WE DO

Agents Are One Part. We Rebuild the Whole Operation.

We don't drop a chatbot into a broken workflow. We find the opportunity, re-engineer the process, and rebuild it with AI as a first-class part of how the work gets done — then tune it until it outperforms.

01

Find the Opportunities

We map your operation and find where AI changes the economics — not where it's trendy, where it pays.

02

Re-Engineer the Process

We rebuild what's being done — AI is a component of a larger operational change, never a bolt-on.

03

Agentic Workflows, First-Class

Agents that act, integrate, and work alongside your team — not chatbots, not copilots.

04

Custom & Tuned Models

Models trained on your data, in your style, for the decisions only your operation has to make.

05

System & Skill Tuning

Proprietary methods that lift the output of skills, tools, and prompts by an order of magnitude.

06

Self-Improving Systems

Systems that adapt and compound — every decision becomes data that sharpens the next one.

We Don't Bolt AI On. We Re-Engineer the Operation.

Most AI projects attach a tool to a process that was already broken and call it transformation. We rebuild what's being done — with agents, tuned models, and self-improving systems as components of a larger change. The result isn't a faster version of the old way. It's a better operation.

Faster Decision Cycles

Multi-system coordination that took days now resolves in minutes. Exceptions get handled before they compound.

Scale Without Adding Heads

Operational output increases without proportional headcount. Your team focuses on judgment, not coordination.

Decisions That Compound

Every decision the system makes becomes data that improves the next one. The longer it runs, the better it gets.

THE COST OF WAITING

Every Quarter You Run It Manually, the Gap Widens

You have the data. You have the processes. You have the systems. What's missing is the team that can re-engineer how it runs and ship it in weeks. The operators who move first compound a structural advantage — in speed, margin, and how much one person can run — that's hard to close later.

02 // HOW IT WORKS

Four Steps to Production

01

Discover

We embed with your team and map how your operation actually runs — every handoff, exception, and decision point.

02

Re-Engineer

We redesign the process and architect the system — integration points, decision boundaries, escalation paths, and human controls — before writing code.

03

Build & Tune

Production-grade deployment with custom-tuned models, full eval suites, monitoring, and version control. Tested against your real data.

04

Compound

Go-live is the beginning. The system feeds outcomes back as signal, sharpens its own logic, and gets better the longer it runs.

03 // WHERE AGENTS DEPLOY

Agents Across Your Operation

Once the process is re-engineered, agents run inside it — deployed across manufacturing and supply chain operations in companies from $50M to $500M+ in revenue.

Production Planning

Agents monitor capacity, flag schedule conflicts, and help planners make faster decisions when conditions change — without waiting for someone to pull the report.

Supply Chain & Procurement

Agents track supplier commitments, surface delivery risks before they become line stoppages, and help purchasing teams stay ahead of disruptions instead of reacting to them.

Cost Control

Agents watch spend against budget in real time, flag variances when they're still actionable, and help operations leaders see where margin is leaking before it's already gone.

Quality & Compliance

Agents support inspection workflows, flag anomalies against your spec tolerances, and maintain the audit trail your team needs without adding documentation burden.

Demand Forecasting

Agents synthesize order data, market signals, and historical patterns to give planning teams a sharper view of what's coming — and how much confidence to put in each number.

Maintenance & Reliability

Agents analyze equipment data to flag degradation early, help maintenance teams prioritize work orders, and reduce the unplanned downtime that breaks production schedules.

04 // WHY BLACKARC

Built by Operators, Not Advisors

Between us, we've deployed 150+ AI agents into production — across federal, SLED, and commercial manufacturing and distribution — built to meet federal security requirements. We bridge C-suite operational experience with hands-on AI implementation.

Joshua Schultz

Operations & AI Strategy

Joshua ran companies before he built AI for them. Years of C-suite operations leadership — scaling PE portfolio companies, running manufacturing and international operations — taught him that the biggest profit lever is eliminating operational friction. He now deploys AI across federal, SLED, and commercial environments: agentic workflows, decision factories, and custom model development that deliver measurable operational gains. Every technology decision comes back to one question: how does this improve operational performance and the bottom line?

C-suite operations leadership and PE portfolio company scaling.
150+ AI agents deployed across federal, SLED, and commercial environments.
Model training & fine-tuning, multi-agent orchestration, and decision factories — built for compliance.

Brad Baker

Emerging Technology & Federal

Brad has been at the front edge of technology his entire career — from the early days of BBS networks to pioneering NVIDIA CUDA for the U.S. Navy, from developing autonomous submarines to leading CTG Federal as an early adopter of disruptive technology. He brings deep technical depth and a track record of getting transformative tech into production before the market catches up.

Pioneered NVIDIA CUDA for the U.S. Navy; developed autonomous submarines.
Leadership at CTG Federal — early adoption of disruptive technology.
Decades bridging cutting-edge research and real client value.

Built for Production, Not Demos

Eval Suites

Every agent ships with quantitative evaluation — regression tests, accuracy benchmarks, and failure-mode coverage before anything touches production.

Version Control

Prompt versions, model configs, tool definitions, and guardrails all tracked. Rollback in minutes, not days. Full audit trail for every change.

Security & Access

Role-based access, scoped credentials, input validation, and output filtering. Agents only see and do what they're authorized to.

Observability

Structured logging, trace IDs, latency tracking, cost monitoring, and drift detection. You see exactly what your agents are doing and why.

Continuous Tuning

Production data drives improvement. We monitor, adjust decision logic, and expand scope based on what actually works — not assumptions.

Human-in-the-Loop

Configurable approval gates, escalation paths, and override controls. Agents augment your team — they don't replace judgment.

05 // LOW-RISK START

A First Engagement Built to De-Risk the Decision

You shouldn't have to bet the operation to find out if this works. The first engagement is tightly scoped, fixed in timeline, and validated against your real data before anything goes live.

Scoped & fixed

A tightly scoped first workflow with a clear timeline and deliverables — no open-ended commitment.

See it before go-live

You review the system against your real data before it touches production. No surprises.

Eval-validated

Nothing ships without quantitative accuracy benchmarks, regression tests, and failure-mode coverage.

Reversible by design

Version-controlled configs with rollback in minutes, human override on every agent, and full audit trail.

Works With Your Stack

SAP · Oracle · Epicor · Infor · Acumatica · Visual · NetSuite · SYSPRO · Plex · IQMS · Global Shop Solutions · Sage — plus any MES, WMS, CMMS, or QMS with an API.

"Your business is the sum of every decision made within it. We help you make better decisions, faster — and build systems that compound."

— BlackArc Industrial

Book a Call

A short call to understand your operation, find where AI changes the economics, and scope a low-risk first engagement. No pitch deck. No pressure.

Book a Call

Read Our Thinking

How we think about AI in manufacturing — small tuned models, agentic workflows, and what production actually requires.

View Insights