Skip to content
Build n Bloom

Sollerius

Bring reviewers the work behind the draft.

Sollerius is the platform we are developing to keep evidence, methods, drafts and review together. The signed-in platform is not complete or available as a hosted service. These are fictional examples captured from local development builds.

Sollerius analysis screen for Northline Studio, fictional. Sample and inference limits, a table of three candidate themes with their source files and AI-proposed challenges, and a side panel of exact source passages from the project lead, delivery consultant and client sponsor. Opens the full capture in a new tab.
T2 interpretation · detail from the saved interface
Sollerius analysis screen for Northline Studio, fictional. Sample and inference limits, a table of three candidate themes with their source files and AI-proposed challenges, and a side panel of exact source passages from the project lead, delivery consultant and client sponsor. Opens the full capture in a new tab.
The selected source passage · same saved interface

Northline Studio · Fictional local build. Focused views of a candidate interpretation and its source. Source counts do not establish prevalence.

Full capture

Start with the engagement.

Keep the brief, sources, working papers and review together. These are actual screens from the local development build.

Actual local Sollerius setup for fictional Cedar Review, saved setup v1 and scope v1. Discovery and preparation show no recorded gap; planning has three items to check and handover has four. Entry and mandate, people, and team and delivery-plan sections are visible.

The engagement starts with a saved brief. Missing planning and handover items remain visible.

Review Practice · Fictional local development example · 25 October 2026

The captures show saved setup, catalogue navigation and a source-change hold. Sollerius is in development; this is not a hosted service or a live connection to external tools.

Make the analysis traceable.

Prepare analysis from the files the engagement is allowed to use, following the method your firm chooses. Keep each interpretation beside the passage behind it, so a consultant can question the reading before it becomes advice.

Source passage

Evidence — 02-client-sponsor.txt: I could not tell which version the team was using.

Interpretation

The sponsor’s version uncertainty and the lead’s report of complete handovers concern distinct aspects of the process. The lead passage challenges a missing-information explanation for two cases; it does not contradict the sponsor’s experience or the distinction made in this theme. Treat its relationship as qualified rather than as an opposing account.

Reason and limits

Reason and limits: Fresh method2 run still uses the same lead quotation as both support and counter-evidence. Human review distinguishes the claim being challenged from the theme itself. This is a new run-specific judgement, not an automatic carry-over of the earlier T1 correction. Three authored accounts establish no frequency, cause or demand.

Text from the saved report: Northline, fictional · Report 12 · source 2. Exact quotations do not establish a valid interpretation. This saved report precedes later scope-change holds.Download DOCXthemes.docx · as saved
Sollerius Knowledge, Methods: Planning commitments and effort, a fictional firm method at approved v2, with five method criteria and a note that firm approval is not expert, buyer or client acceptance. Opens the full capture in a new tab.

The firm’s method lives in its own register, versioned and approved separately from any report. Fictional firm method, approved v2; firm approval is not expert, buyer or client acceptance.

Full capture

Use the firm’s method because it fits this question. Keep the reason with the work, so the next reviewer can follow the choice.

Northline · fictionalCopy 83f3a5f0TXT

A precedent selected for the actual question

The permitted precedent set is limited to owned fictional firm methods; the knowledge search returned no other permitted entries. The closest fit is the Cross-case interview synthesis method, version 2. It fits because the present question is not whether a service should be launched, but how to compare selected role accounts while keeping a reported experience separate from a causal explanation. It transfers four usable disciplines: quote-linked candidate themes; visible source roles; a specific explanation of what any apparently contrary account challenges; and an explicit limit on prevalence, causation, demand and value claims.

1 more saved passage in the file.

Review the version you will stand behind.

Read the draft, test the interpretation and record why it is ready—or what needs to change. The review belongs to that saved version. A changed source puts it back in review; it does not inherit the earlier decision.

Phone view of the Harbour Field Services review, fictional: report v7 on source v1, Accepted for this saved version, with the reason, which states that this operator check is not qualified expert acceptance, and the intended use. Opens the full phone capture in a new tab.

Accepted for this saved version

Harbour · fictionalReport 7 · source 1DOCX

Reviewed operational diagnostic

For 26 October 2026, 300 available hours at an estimated five hours per job model 60 jobs; this is not measured throughput. The confirmed-only comparison uses 60 jobs. The planning register still contains 80 jobs, including 20 provisional jobs dependent on client access. If none of those provisional jobs release during the week, the confirmed-only comparison remains 60; if all release, the potential workload is 80 against the same 60-job model. The planning register remains 80 in both scenarios. These conditional comparisons do not establish delivery readiness, spare capacity, staffing need or a commitment. Confirm actual access-release dates, validate job mix and measured effort, and compare scheduling and service scope before deciding.

The effort input is an estimate. Validate the period, job mix and available options before any commitment. This comparison does not establish staffing viability or guaranteed delivery.

An operator check in a local build, not qualified expert or client acceptance. The phone capture shows the decision and its reason; the reader shows the saved report text.

Full desktop capture
The consultant prepares
Tests the interpretation against the source passages and brings the qualification into the wording.
The reviewer decides
Approves, returns or holds the exact saved version, and records why.
The client lead issues
Client issue, client acceptance and billing stay separate from the internal review.

A correction, a changed source, then a hold.

Fictional local study · Northline only

These saved Northline materials make one bounded point: a person corrects the meaning; a later source change reopens review; an AI draft cannot clear what remains unresolved. They are not an end-to-end approved handover, and nothing shown here was issued.

  1. 01 · Correct the reading

    Maya corrects the relationship between the sponsor and lead passages in Report 12 · source 2. The report records a human semantic correction, not an automatic conclusion.

  2. 02 · Hold the changed basis

    Later, Report 13 · source 3 needs method and applicability review. It does not inherit any earlier decision.

  3. 03 · Keep handover held

    Draft a2c34f25 records the unresolved hold. Maya clarifies it; the draft remains unreviewed and cannot establish closeout, issue or acceptance.

Carry the decision into the next job.

A reviewed report is part of a larger engagement. Handover needs the scope, delivery planning needs the effort assumptions, and later work needs to know what may be reused. Your team decides what carries forward.

Sollerius delivery plan for Birchwell Partners, fictional. Role constraints for plan v3 compare analyst and reviewer hours; role-limited case equivalents fall from 3 baseline to 1.5 revised, and neither comparison is an actual allocation, completed engagement or payroll saving. Opens the full capture in a new tab.

Plan around the reviewer, too.

Birchwell’s plan v3 compares analyst and reviewer hours separately. Neither comparison is an actual allocation, a completed engagement or a payroll saving.

Sollerius handover tab for Birchwell Partners, fictional. Internal proposal v2, What delivery is being asked to take on, listing the decision, deliverable and exclusions. Opens the full capture in a new tab.

Hand the scope to delivery.

Internal proposal v2 carries the decision, deliverable and exclusions across. It is internal planning, not a client win or permission to start.

Sollerius engagement learning for Willowmere Practice Lab, fictional. The Coaching and practice tab is Reviewed for internal use; the working copy states it is fictional practice material, not an account of any employee, client or real event. Opens the full capture in a new tab.

Turn a correction into practice.

Willowmere’s coaching exercise is reviewed for internal use. It is fictional practice material, not an account of any employee, client or real event.

Every job here is a design in the library, with its inputs, output and human decision. 41 workflows and 16 agent responsibilities.

Capabilities and scope.

All 36 capabilities in seven groups. Each one states what it helps your team do, its current scope and the evidence behind it.

Engagement workspace6 capabilities

Keep the original question in view as the work moves between people.

Start the next engagement with its own workspace.

Give each engagement its own scope, sources and saved work. Start cleanly without copying another client’s material into the next job.

Demonstrated locally. No multi-tenant production claim.

Evidence and current scope

v37/v43/v47/v48; authenticated grants tested separately.

See what is waiting, who owns it and what it affects.

Keep prerequisites, named owners and dates together. Distinguish an internal target from a client promise and show which later work depends on it.

Demonstrated locally. Broader plan is partial.

Evidence and current scope

v40/v44/v45; internal targets, client promises and billing conditions stay separate.

Turn a client instruction into clear follow-through.

Record the client instruction, its conditions and the follow-through it needs. Keep it distinct from an internal review or an assertion of client acceptance.

Demonstrated locally. Not automated communication.

Evidence and current scope

v45 local simulation; no actual client message, consent or external acceptance is inferred.

Connect what you pursue with what you can deliver.

Prepare discovery, proposal scope and fee assumptions from the actual service and its delivery needs. Carry the internal decision into the handover.

Local opportunity-to-handover preparation demonstrated. This does not establish signed terms, a client win or an external CRM update.

Evidence and current scope

v43 W01–W06; no external CRM write, signed contract or sales outcome inferred.

Evidence and analysis5 capabilities

Use evidence and a chosen method to prepare analysis, then test what it supports.

Prepare analysis from the evidence and method you provide.

Prepare a substantive working analysis using the brief, selected permitted evidence and a chosen firm method. Your team tests the interpretation before relying on it.

Demonstrated locally. Your team remains responsible for interpretation, quality and completeness.

Evidence and current scope

Capacity and qualitative paths received; provenance does not establish semantic correctness.

See alternatives, gaps and unanswered questions.

Bring possible rival explanations, missing evidence and uncertainty into view, so the reviewer can challenge the first account rather than simply polish it.

Demonstrated locally. Does not guarantee completeness.

Evidence and current scope

v42/v46; exact-quote and semantic-relationship checks are distinct.

See the assumptions behind the scenario.

Calculate supported comparisons separately from generated prose. Inspect the labelled inputs and assumptions before using a scenario in a decision.

Supported calculations demonstrated locally; wider operating work is partial. Supported calculations only.

Evidence and current scope

v36/v43/v44/v45; new v49 cash/licence calculations have targeted checks, whole operating workflows paused.

Import evidence with a record of what arrived.

Bring permitted CSV, text or Markdown evidence into the engagement with a record of its purpose, version and contents.

Demonstrated locally. CSV/TXT/MD only; no PDF/DOCX/XLSX import.

Evidence and current scope

CSV/TXT/MD case import received; private v49 firm-input two-file admission also recorded. Native PDF/DOCX/XLSX import and live cloud-file connectors are not established.

Check imported records and resolve exceptions.

Check incoming records for missing fields, duplicates and inconsistent identities. Resolve exceptions without silently overwriting a valid earlier source.

Demonstrated locally. Local scope.

Evidence and current scope

Quoted CSV and BOM support repaired in v48; exact reversals retain lineage.

Report review and outputs5 capabilities

Correct the wording, review the exact version and take an ordinary editable deliverable forward.

Download the reviewed work as an editable Word document.

Take the exact reviewed wording into an editable Word document, with its evidence and review context available for the next person.

Demonstrated locally. Editable Word output is supported; native Word import is not.

Evidence and current scope

v38/v39/v41/v46; Word export does not imply native Word import or client issuance.

Keep a readable copy of the reviewed work.

Keep a readable HTML copy of the reviewed report and inspect earlier versions when you need to understand what changed.

Demonstrated locally. Readable HTML reports and version history; PDF output and a full document store are not established.

Evidence and current scope

Local report paths received; this does not establish PDF export or publication.

Refresh the assessment when the evidence changes.

Identify work whose evidence, method or scope has changed. Prepare a new version for review while keeping the earlier wording and decisions available.

Demonstrated locally. Human reassessment required.

Evidence and current scope

Restoring identical source bytes does not silently revive approval.

Firm methods and knowledge4 capabilities

Keep the reason behind a correction. Decide which methods and learning the firm can reuse.

Put your firm’s methods to work.

Use an owned, versioned method in the engagement. Explain why it fits this question and keep the method owner’s approval separate from the report review.

Demonstrated locally. Human ownership and quality remain required.

Evidence and current scope

v41/v46; firm approval and engagement applicability have separate versions.

Keep the reason behind a professional correction.

Record what the interpretation got wrong and why the professional correction matters. Use selected examples to prepare coaching or propose a method improvement.

Demonstrated locally. Method promotion requires approval.

Evidence and current scope

v46/v47; no employee scoring, automated competence assessment or proven training outcome.

Keep firm methods and client evidence in their proper context.

Keep private case material separate from selected reusable text. Record its origin, owner and permissions before making it available beyond the case.

Demonstrated locally. Permissions must be configured.

Evidence and current scope

v47; withdrawal blocks future availability/export but cannot recall prior downloads; rights are user-recorded.

Improve the working method from real feedback.

Carry an observed correction into an owned method revision, then use the approved revision in a fresh preparation. Review reusable learning separately.

Local learning mechanisms demonstrated. Method revisions and reuse require human approval; model training is not automatic.

Evidence and current scope

v46/v47; no automatic model training or automatic method approval.

Capacity and service economics5 capabilities

Include reviewer effort, delivery dependencies and commercial conditions in the comparison.

Plan around the roles the work actually needs.

Compare analyst and reviewer effort separately. See when a different role becomes the constraint instead of treating all available hours as interchangeable.

Core demonstrated locally; broader review is partial. Scenario, not capacity guarantee.

Evidence and current scope

v44 core received; W20 whole capacity/pipeline review remains paused.

Connect scope changes with delivery planning.

Identify when a scope change makes the resource plan out of date. Keep the earlier assumptions and prepare a revised plan before making another commitment.

Core demonstrated locally; broader review is partial. Partial professional job support.

Evidence and current scope

v44; do not silently change the recorded proposal fee or carry actuals into a new plan.

Compare the full effort and cost of a service.

Include preparation, review, correction and other known costs in the service comparison. Keep unknown costs visible rather than presenting an incomplete subtotal as a final margin.

Local scenarios demonstrated. No savings, margin or profitability promise.

Evidence and current scope

v44/v48; cost scenarios are not measured savings, fees, margin guarantees or cash.

Keep invoices, receipts and outstanding balances distinct.

Compare imported invoices with matched receipts, outstanding balances and reversals. Keep a recorded payment assertion distinct from verified cash.

Demonstrated locally with imported records. Live accounting connections and actual cash require separate verification.

Evidence and current scope

v36/v45; imported accounting assertions are separate from live accounting-provider actions and cash.

Implementation and service development7 capabilities

Shape the first installation around your services, with clear human ownership and room to investigate new ones.

Explore a service your clients could use.

Bring the proposed client decision, evidence needs, delivery assumptions and a competent current alternative into a service pilot decision pack.

Preparation demonstrated locally. Demand and launch unproven.

Evidence and current scope

v48; positive proactive investigation works without an inbound enquiry; real usefulness/demand/payment/renewal remain unverified.

Shape the platform around how your firm works.

Combine a shared working structure with your firm’s methods, selected services and ownership. Scope the installation around the work your team needs to do.

Configuration demonstrated locally. Bespoke scope is explicit.

Evidence and current scope

Multiple authored cases tested; this is not proof of production multi-tenant service.

A working environment organised around the engagement.

Move between the queue, evidence, working papers, comparisons and operating records in a layout suited to the task.

Current interface checked locally. Practitioner acceptance and a complete accessibility audit remain outstanding.

Evidence and current scope

v50 Opus-guided desktop/intermediate/narrow receiving, bounded keyboard/contrast checks and registry/export preservation. Actual200% zoom, complete screen-reader audit and practitioner acceptance remain unverified.

Use AI within the work, with visible inputs and results.

Open a supported job in its engagement context and inspect the inputs, preparation and result. Keep the next human decision visible with the work.

27 local paths checked; 14 paused. Eight role configurations are inspectable; these are not installed autonomous workers.

Evidence and current scope

41 workflow definitions;8 inspectable role configurations;16 responsibility specifications in the research catalogue. None of those counts establishes deployed autonomous workers.

Start with work worth improving—and keep what works.

Compare a bounded improvement with the competent process or tool your team already uses. A useful experiment, a simpler change or no build can each be a sound decision.

Local decision mechanisms demonstrated. Practical usefulness needs assessment with your team.

Evidence and current scope

v42/v48; practical usefulness and buyer acceptance require later evidence.

Build the capability with your team.

Make setup, acceptance, handover, remaining obligations and support ownership explicit with your team. Agree the commercial scope for the installation.

Support preparation demonstrated locally. Installation, support responsibilities and any continuing arrangement are agreed separately.

Evidence and current scope

v45; no provider access change, legal discharge, support SLA or automatic retainer.

Discuss continuing access shaped around your firm’s work.

Discuss the access, onboarding, supported use and responsibilities your firm would need. A separate agreement would define continuing platform access.

Proposed. Continuing access, price, support, renewal and contract terms are not established.

Evidence and current scope

No validated platform price, subscription billing, access contract, paid demand or renewal established.

Access and working records4 capabilities

Give people the right responsibilities and preserve the history needed to resume or reconcile work.

Set access and review responsibilities for a workspace.

Give owners, leads, reviewers, finance staff and viewers distinct responsibilities, with workspace and engagement access chosen for the work.

Access controls tested locally. Hosted tenancy, SSO and security certification are not established.

Evidence and current scope

v38 and later targeted grant checks; production tenancy, SSO and hosting are not established.

Keep work and its history available across restarts.

Resume supported local work after a restart with its saved records and histories, rather than reconstructing the last working state.

Demonstrated locally. Production recovery arrangements remain to be established.

Evidence and current scope

v36–v48 receiving; local persistence is not production disaster recovery.

Restore working records, then check their review state.

Preview and restore the supported workspace records, then reassess their review state before taking work forward.

Bounded workspace recovery demonstrated locally. The JSON backup excludes separate AI and backing stores.

Evidence and current scope

Backup excludes separate immutable AI/backing stores; complete portable server recovery is not established.

Resolve conflicting changes before replacing valid work.

Surface conflicting changes instead of silently replacing valid work. Confirm what was saved before the next review or download.

Demonstrated locally. Production concurrency and load have not been established.

Evidence and current scope

Bounded CAS/retry and save-order tests; no production concurrency/load claim.

Where Sollerius stands.

As of 25 October 2026, Sollerius is a local development build. The examples are actual exports from fictional sources, with AI preparation and human review tested locally. They are not client results or hosted availability. Microsoft examples are not connected; a Word export is an editable file, not a Microsoft integration.

Technical inventory

Saved outputs

The saved examples demonstrate local AI preparation, human corrections, review and editable outputs using fictional sources, from separate engagements. The dated v50 catalogue records 27 of 41 locally checked paths, 14 paused paths and eight inspectable role configurations. Its 16 responsibility specifications and 16 selected exports are different views of that work, not counts of installed workers.

Still to be established

Professional quality, buyer acceptance, business outcomes and production readiness are not established by these examples. The connection directory contains one local import/output route and 14 external examples or proposals. Microsoft, Google, CRM and accounting examples are not live connections. Hosted operation and professional usefulness require their own evidence.

Each installation

Source rights, connections, supported work, onboarding, access and continuing responsibility are agreed for each installation. The quote defines acceptance tests and the applicable service guarantee. Thirty calendar days of in-scope stabilisation follow written acceptance; later support is separately agreed. Continuing platform access, its price and contract remain proposed.

You can commission implementation while Sollerius is in development. We scope a build around your service, tools and review process, then agree how your team will test it. Start with a free Fit Call; a clear scope can be quoted directly.

Discuss your work How a build begins