RECURSIVEINTELL

01SERVICES + CONSULTING

Choose the smallest
useful intervention.

Start with a fixed, inspectable scope. Every offer names the required inputs, retained artifact, explicit boundary, acceptance test, and next decision before implementation expands.

01Focused consulting

AI Workflow Map

Best fit
You know a process wastes attention, but the right automation boundary is still unclear.
Inputs
One repeated workflow, its users, current tools, representative inputs, and known exceptions.
Deliverable
A current-state map, automation candidates, authority boundaries, risk register, and smallest useful pilot recommendation.
Explicit boundary
Analysis and design only. No production integration or guarantee of savings.
Acceptance test
The map names owners, inputs, decisions, exceptions, proposed effects, and a falsifiable next gate.
Ask about this offer →
02Bounded implementation

Custom Agent Pilot

Best fit
One agent-assisted workflow has a clear owner and can be tested without reorganizing the whole business.
Inputs
The approved workflow map, sample tasks, authorized tools or data, and named approval points.
Deliverable
A working pilot, configuration, test fixtures, explicit failure behavior, operator walkthrough, and handoff notes.
Explicit boundary
One agreed workflow and integration surface. External service fees and later expansion remain separate.
Acceptance test
The pilot completes the agreed examples, refuses or escalates boundary cases, and leaves an inspectable execution record.
Ask about this offer →
03Source-aware system

Business Knowledge Build

Best fit
Useful answers are trapped across documents, notes, policies, or specialist knowledge that people repeatedly search for.
Inputs
An authorized source set, access rules, representative questions, update process, and escalation owner.
Deliverable
An indexed knowledge path, source-aware answer interface, update procedure, failure states, and evaluation set.
Explicit boundary
The system surfaces evidence; it does not replace policy owners or guarantee factual correctness.
Acceptance test
The agreed questions return relevant sources, disclose missing evidence, and preserve the defined access boundary.
Ask about this offer →
04Technical advisory or operations

Systems Consulting + Ongoing Care

Best fit
You need architecture judgment, reliability work, local-first deployment, Hermes customization, or continued improvement after a pilot.
Inputs
The exact system boundary, current source or configuration, desired decision, and evidence required at the end.
Deliverable
A scoped review, implementation pass, reliability audit, or agreed maintenance cadence with written receipts and handoff.
Explicit boundary
Not a penetration test, certification, compliance opinion, or unlimited support agreement.
Acceptance test
The engagement closes its named decision or change with mapped checks, results, limitations, and remaining delta.
Ask about this offer →

FOCUSED TECHNICAL CONSULTING

Use the depth
without buying a platform.

Consulting can end in a decision, architecture, audit, implementation plan, proof harness, or bounded source change. The useful output is an artifact your team can evaluate and retain.

01

Agent runtime architecture

Graph design, model and tool boundaries, memory routes, failure states, receipts, and operator control.

02

Hermes integration

Skills, MCP servers, local memory, Rust-backed services, multi-agent workflows, and bounded performance work.

03

Local-first AI

Deployment topology, data residency, provider boundaries, offline behavior, recovery, and explicit degraded modes.

04

Reliability + proof

Hostile workflow review, replay cases, claim/evidence boundaries, acceptance tests, and release gates.

ENGAGEMENT CONTRACT

Clear before
the first build.

Pricing: Fixed scope confirmed after fit review. No project begins from a vague estimate disguised as a commitment.

External costs: Model/API, hosting, phone, email, and other third-party service fees stay visible and are assigned in the written scope.

Ownership: Source, credentials, operational responsibilities, support period, and handoff artifacts are named before implementation.

Claims: No guarantee of savings, accuracy, revenue, security, compliance, or universal automation fitness.

FIT REVIEW

Send the system,
constraint, and
desired evidence.

You do not need a polished brief. Four concrete lines are enough to decide whether a useful engagement exists.

Describe the work →Inspect public work