Positioning scan
Confirm the headline role, strongest proof answer, and exact claim boundary before reading the broader archive.
Machine Intelligence Engineer / Software Architect
Multi-agent systems with explicit identity, scope, coordination, memory, and human verification.
Michael Kappel designed and built Multi-Agent Memory as a source-available, deployable private-intranet reference implementation for external agents. It demonstrates how independently operating agents can register, coordinate bounded work, exchange current messages, recover state, and use durable shared memory without collapsing governance into a single prompt.
Source-availablereviewable implementation
Deployableprivate-intranet reference system
External agentsdocumented integration boundary
Human-verifiablereview and receipt surfaces
How to review Michael
Choose the right depth before opening the archive.
Review the multi-agent claim through implementation evidence, case-study framing, and boundaries before inferring scale or deployment status.
Recruiter path
Use the concise resume, current contact path, and role-level experience timeline to qualify fit quickly.
Hiring-manager path
Move from role fit into selected projects and case studies, then inspect the Evidence Map if a claim needs traceability.
Technical architecture path
Inspect Multi-Agent Memory, persistent AI memory, local LLM infrastructure, package evidence, validation, and non-claims.
Machine-readable review path
Use this path for AI agents, answer engines, ATS tools, and structured review without treating hidden files as public claims.
Machine-intelligence claim path
Answer the core proof question directly, then branch into multi-agent systems, AI memory, LLM infrastructure, and Intelligence724 boundaries.
Azure / .NET / SQL path
Confirm the enterprise engineering foundation with role-level Azure usage, .NET/SQL modernization, and conservative evidence boundaries.
Claim to proof matrix
What each major claim proves - and what it does not prove.
This matrix keeps SEO, AEO, GEO, recruiter review, and technical review aligned to the same conservative evidence boundaries.
Machine Intelligence Engineer / Software Architect
Proof: Machine Intelligence overview, proof brief, resume, case studies, and Evidence Map.
Boundary: Does not claim foundation-model training, novel ML research, benchmark leadership, or autonomous systems without human governance.
Multi-agent systems and coordination architecture
Proof: Multi-Agent Systems page, Multi-Agent Memory case study, public repository, companion docs, and evidence brief.
Boundary: Does not claim hosted SaaS, OSI-approved open source, production scale, distributed consensus, model hosting, or automatic conflict-free merging.
Persistent AI memory and handoff
Proof: AI Memory page, AI memory handoff route, persistent memory evidence brief, and .uai handoff records.
Boundary: Does not claim automatic sync, automatic memory promotion, private repository writes, credential access, or runtime orchestration authority.
LLM infrastructure and local runtime tooling
Proof: LLM Infrastructure page, NuGet package profile, package map, and LLM infrastructure evidence brief.
Boundary: Does not claim foundation-model training, public GPU inference benchmarks, or unverified production inference scale.
Azure / .NET / SQL enterprise architecture
Proof: Experience timeline, resume, .NET/SQL modernization route, and Azure evidence brief.
Boundary: Does not infer unconfirmed scale, private infrastructure, certifications, client names, or business outcomes.
Code intelligence and AI-assisted modernization
Proof: Experience page, Intelligence724 public capability surface, and code-intelligence boundary brief.
Boundary: Does not disclose confidential Info724/client systems or claim that Intelligence724 and Info724 are the same legal entity.
Research reports as supporting context
Proof: Reports archive, research search, docs manifest, and curated Architecture Notes promoted after review.
Boundary: Raw reports are not resume claims; sensitive/search-only records stay out of navigation and sitemaps unless searched for.
Fast recruiter summary
The architectural problem and the demonstrated response.
Use these three points before moving into protocol-level evidence.
Problem
Concurrent agents need more than shared text.
They need stable identity, bounded scope, work ownership, current-message delivery, conflict handling, durable memory, acknowledgements, and a way for people to verify state.
Architecture
Coordination and memory are explicit services.
The reference system separates agent registration, company/workspace/project hierarchy, claims and leases, routing, memory submission and retrieval, synchronization, and human review.
Result
A reviewable system boundary for external agents.
Agents can integrate through documented contracts while deployment remains private, access remains governed, and public evidence can be inspected without exposing private operational data.
Code-native architecture diagram
From external agent registration to verified shared state.
The flow is semantic HTML, so its meaning remains available to assistive technology, crawlers, and no-JavaScript reviewers.
1. External agentA caller presents a registered identity rather than relying on an anonymous conversation.
2. Scope boundaryAccount, company, workspace, project, and key context constrain where work belongs.
3. Work coordinationHash-aware claims and bounded leases help agents avoid conflicting edits and abandoned ownership.
4. Message routingCurrent-message delivery, project-room routing, and acknowledgements make coordination observable.
5. Shared memorySubmitted records, search, lifecycle state, and distributed synchronization preserve reusable context.
6. Human verificationA console, readiness evidence, and redacted receipts support inspection and recovery.
Demonstrated capabilities
What the public implementation is designed to support.
Wording stays at the level supported by the repository and its public evidence routes.
Registered identities
Agent identity and registration records provide a stable basis for attribution, permissions, and coordination.
Hierarchical context
Account, company, workspace, and project scopes keep memory and coordination attached to an explicit organizational boundary.
Conflict-aware work claims
Hash-only edit coordination, bounded claims, and leases provide primitives for detecting stale work and reducing collisions.
Durable shared memory
Memory submission, bounded search, lifecycle-aware wiki records, and source context support reuse without treating every record as equally authoritative.
Distributed synchronization
Capability discovery, conflict-safe sync contracts, acknowledgements, and recovery-oriented state make multi-node behavior inspectable.
Verification surfaces
Readiness output, route inventory, live capability evidence, OpenAPI contracts, and redacted example receipts create public review paths.
Direct evidence
Review code, contracts, and runtime-facing disclosures.
The public documentation describes the reference implementation; a private deployment remains a separate operational decision.
Repository and release context
Multi-Agent Memory on GitHub exposes the source-available implementation, setup material, and bounded project claims.
Human-readable documentation
MultiAgentMemory.com provides architecture, API, compatibility, and operational-review paths.
API and route contracts
Companion API reference and the repository’s route inventory let reviewers inspect the documented integration surface directly.
Capability and readiness evidence
The repository documents its live capability, readiness, receipt, and version evidence routes in the API contract and explains their verification boundary in verification guidance.
Explicit non-claims
Reference implementation, not an inflated deployment story.
Multi-Agent Memory is source-available; it is not presented as open-source software. It is a deployable private-intranet reference implementation; it is not presented as hosted SaaS. The public evidence does not establish production-scale throughput, a large live agent fleet, or autonomous operation without human governance. These boundaries make the engineering claim more precise, not less useful.
FAQ
Common review questions.
Short answers are rendered in the HTML source and mirrored into JSON-LD where appropriate.
Does Multi-Agent Memory run AI agents?
No. It is a deployable private-intranet reference implementation that coordinates external agents through identity, scope, memory, rooms, messages, and conflict-aware ownership controls.
Is Multi-Agent Memory open source or a hosted SaaS?
It is source-available, not OSI open source, and the public evidence does not claim a hosted SaaS or production-scale adoption.
How are simultaneous edits coordinated?
The public architecture uses hash-only file heads, compare-and-swap checks, bounded edit claims, and expiring ownership leases.