Blog

Applied AI, with no shortcuts in strategy.

Articles that help leaders prioritize, structure and monitor Artificial Intelligence initiatives.

Cyan and violet workflow ribbons crossing an amber fracture under a transparent forensic lens that reveals evidence and reconstructs the path
Hybrid squads · 7 min

Hybrid squads: how to investigate failures across people and AI agents

When work passes through people, agents and systems, the visible error rarely explains the whole incident. Learn how to investigate without losing accountability or learning.

Oct 23, 2026 · NM SolutionsContinue reading ↗
Transparent prism filtering cyan customer signals through evidence layers while an amber unauthorized commitment is diverted to a violet review point
AI agents · 7 min

AI agents in B2B support: how to respond without making unauthorized commitments

A plausible answer can create an invalid deadline, discount or promise. Learn how to give B2B support agents context without transferring commercial authority.

Oct 21, 2026 · NM SolutionsContinue reading ↗
Circular release mechanism with violet, amber and cyan modules inside a transparent chamber for controlled AI agent changes
Operations management · 7 min

Operating AI agents: how to change prompts, models and tools without losing control

A small adjustment can alter thousands of agent decisions. Learn how to version, evaluate, release and reverse changes with operational control.

Oct 19, 2026 · NM SolutionsContinue reading ↗
Two different digital systems exchanging cyan data packets through a transparent violet translation chamber that isolates one changed amber packet
Integrations · 7 min

AI agent integrations: how to keep an API change from stopping operations

A renamed field or changed status can stop an agent workflow. Learn how to protect integrations with explicit contracts, adapters, tests and controlled rollout.

Oct 16, 2026 · NM SolutionsContinue reading ↗
Luminous business requests crossing three transparent approval portals, with violet human decision points and an amber exception diverted to a review chamber
Process automation · 7 min

Approval automation: how to set authority levels without turning control into a bottleneck

Approval workflows become slow when every request follows the same path. Learn how to define authority levels, deadlines, substitutes and evidence without weakening control.

Oct 14, 2026 · NM SolutionsContinue reading ↗
Floating stepped platforms distributing luminous work capsules between cyan AI tokens and violet human judgment tokens, with an amber congestion cluster moving toward open capacity
Hybrid squads · 7 min

Hybrid squads: how to organize queues and priorities between people and AI agents

Adding agents can accelerate one stage and overload the next. Learn how to define queues, work limits, priorities and handoffs for a hybrid squad.

Oct 12, 2026 · NM SolutionsContinue reading ↗
Three translucent supplier proposal sets being aligned by a cyan comparison prism while an amber unresolved item remains separate from a violet approval marker
AI agents · 7 min

AI agents in procurement: how to compare suppliers without outsourcing the decision

An AI agent can organize proposals and reveal differences, but it should not turn incomplete documents into an automatic supplier decision. Learn how to design a controlled comparison.

Oct 9, 2026 · NM SolutionsContinue reading ↗
Modular AI workflow rerouting work around an isolated failed component through fallback, queue and manual recovery paths
Operations management · 7 min

Operational continuity for AI agents: what to do when a model, API or data source fails

An AI agent can depend on models, APIs and data sources that will eventually fail. Learn how to plan fallback, queues, human recovery and safe return to normal operations.

Oct 7, 2026 · NM SolutionsContinue reading ↗
Digital agent flow crossing only authorized translucent access gates toward separate business systems while blocked routes remain closed
Integrations · 7 min

AI agent integrations: how to grant access without creating a superuser account

Connecting an AI agent to CRM, email and ERP with broad credentials creates avoidable risk. Learn how to design access around tasks, resources and business limits.

Oct 5, 2026 · NM SolutionsContinue reading ↗
Hybrid operations squad with four connected stations representing process ownership, an AI agent, human review and platform support
Hybrid squads · 7 min

Hybrid squads in practice: how to divide responsibilities between people and AI agents

When people and AI agents share a workflow, unclear ownership becomes an operational risk. Learn how to define roles, handoffs, approvals and escalation criteria.

Oct 2, 2026 · NM SolutionsContinue reading ↗
Abstract process map with one carefully selected luminous workflow and secondary routes containing branches, returns and exceptions
Process automation · 7 min

Process automation: how to choose the first workflow without increasing rework

High volume does not always mean a good first automation. Learn how to assess repetition, exceptions, data, risk and business outcomes before building.

Sep 30, 2026 · NM SolutionsContinue reading ↗
Abstract AI operations center with a luminous core and five signal paths converging on a monitoring surface
Operations management · 6 min

AI agents in production: five metrics to track before expanding autonomy

An agent can complete many tasks and still create rework. Learn which operational signals help leaders decide when to maintain, restrict or expand autonomy.

Sep 28, 2026 · NM SolutionsContinue reading ↗
Illustration of two systems connected by a luminous bridge, with one order in transit and a copy set aside for review
Automation and integrations · 6 min

CRM and ERP automation: how to avoid duplicate orders and conflicting data

Connecting systems is not enough. Learn how to define the source of data, recognize repeated attempts and track exceptions without turning automation into rework.

Sep 25, 2026 · NM SolutionsContinue reading ↗
AI agents · 6 min

AI agents: what to automate and when to require human approval

A practical approach to separating suggestions, reversible actions and sensitive decisions, with clear ownership, approval criteria and quality metrics.

Sep 23, 2026 · NM SolutionsContinue reading ↗
AI Squads · 6 min

AI Squads: from strategy to operational results

Why multidisciplinary teams are essential to connect technology, business, data and people and turn AI into measurable results.

Aug 25, 2026 · NM Solutions
Hybrid squads · 7 min

Hybrid squads: people, AI agents and robots working together

Learn how people, AI agents and automation robots can operate as one multidisciplinary team, with clearly defined roles, boundaries and human oversight.

Sep 4, 2026 · NM Solutions