Ask the business
Plain-language questions over ERP, CRM and reports. Numbers come with the records they were built from.
Systems
- ERP
- CRM
- BI
- Spreadsheets
MetricHours → minutes per question
Specific problems, not AI adoption
A library of the problems we solve most often, by function and by industry: what the platform does, which systems it connects, and the metric we use to measure the result. Any of them can become a pilot. If yours isn't on the list, that doesn't mean we don't solve it.
By function
What changes in each department
For each function: typical use cases, and the metric we use to measure the result. Details for every use case are in the library below.
Leadership
Plain-language answers about the business, a daily digest of deviations, tracking of directives, meeting preparation.
Metric · time to answerFinance
Counterparty reconciliations, receivables control, consolidated reporting from ERP and accounting, source-document checks, plan vs. actual.
Metric · days to closeOperations & production
Output plan vs. actual, downtime, quality control, procurement and inventory, on-site video analytics.
Metric · downtime, lossesSales & marketing
Lead qualification, proposal drafting, channel analytics, monitoring of sales activity in the CRM.
Metric · conversion, cycle timeContact center
Multilingual voice and text agents, request routing, response-time tracking, call analysis.
Metric · resolved without an operatorLegal & records
Contract review against a checklist, regulatory search, standard document drafting, deadline control.
Metric · time per contract reviewHR
Résumé screening, policy questions from staff, onboarding, a company knowledge base.
Metric · time to hireIT
A single access point to data for analytics, first-line internal support, systems documentation.
Metric · first-line ticket loadOne data model
Every use case runs on the same model of your organization. The second one ships faster than the first, and each one after that costs less.
How a use case works
Every use case is built the same way
Whether it is a reconciliation, a citizen request or a contract review, the path is the same: your systems feed one model of the organization, an agent does the routine work, a person approves what matters, and the result goes back into the system of record.
- 01
Your systems
ERP · CRM · documents · mail · telephony · cameras
- 02
Organization model
Entities, processes, people and rules, linked
- 03
Agent
Executive · Operations · Documents · Knowledge · Voice · Vision
- 04
Human approval
Required above $500,000 or low confidence
- 05
Action + audit log
Written back, every step traceable
- Answers and actions cite their source records
- Read-only by default; write access is granted per action
- Roles and permissions come from your directory
- Runs in our cloud in your region, your private cloud, on-premises or air-gapped
Use-case library
Twenty-eight use cases, with the systems and the metric
Typical scope for each use case. The systems listed are examples of what we connect; the exact set is confirmed during the diagnostic.
Leadership
MetricTime to answer
Daily deviation digest
Each morning: what moved outside plan, by branch or project, ranked by money at stake.
Systems
- ERP
- Planning files
MetricDeviations seen the same day
Directive tracking
Decisions from meetings become tasks with owners and deadlines. Overdue items surface on their own.
Systems
- Calendar
- Task tracker
- DMS
MetricDirectives closed on time
Meeting preparation
A one-page brief before each meeting: open issues, latest figures, previous decisions.
Systems
- Calendar
- DMS
- CRM
MetricPreparation time
Finance
MetricDays to close
Counterparty reconciliation
Matches statements, invoices and the ledger, drafts reconciliation statements and flags every discrepancy.
Systems
- ERP
- Bank statements
MetricDays per reconciliation
Receivables control
Aging by customer, drafted reminders, escalation of large amounts. Write-offs above $500,000 need a person.
Systems
- ERP
- CRM
MetricOverdue receivables
Consolidated reporting
The monthly pack assembled from several entities and accounting systems, with variances explained.
Systems
- ERP
- Accounting
- Spreadsheets
MetricDays to close
Source-document checks
Invoices and delivery notes checked against contracts and purchase orders before payment.
Systems
- DMS
- ERP
- Scans
MetricErrors caught before payment
Operations & production
MetricDowntime, losses
Plan vs. actual by shift
Output against plan by line and shift, with the likely cause of each gap.
Systems
- MES
- ERP
- Sensor exports
MetricTime to detect a deviation
Procurement and inventory
Stock-outs, excess stock and late suppliers flagged before they stop production.
Systems
- ERP
- WMS
- Supplier mail
MetricStock-out days
Safety and process video
Protective equipment, restricted zones and occupancy watched on your existing cameras.
Systems
- Cameras
- VMS
MetricViolations detected
Engineer's assistant
Process engineers ask about procedures, past incidents and equipment manuals and get quoted answers.
Systems
- DMS
- Maintenance logs
- Manuals
MetricTime to find an answer
Sales & marketing
MetricConversion, cycle time
Lead qualification
Incoming requests scored, enriched and routed to the right person within minutes.
Systems
- CRM
- Web forms
MetricTime to first response
Proposal drafting
Draft proposals from price lists, past proposals and the client's request, ready for a manager to edit.
Systems
- CRM
- DMS
- Price lists
MetricHours per proposal
Pipeline monitoring
Stalled deals, missing next steps and silent accounts flagged to managers each week.
Systems
- CRM
- Calendar
MetricCycle time
Contact center
MetricResolved without an operator
Voice and text agents
Answer routine questions in the customer's language by phone, chat and messengers, around the clock.
Systems
- Telephony
- CRM
- Knowledge base
MetricResolved without an operator
Routing with summaries
Classifies each request, sends it to the right queue and hands the operator a short case summary.
Systems
- Telephony
- Ticketing
- CRM
MetricAverage handling time
Call quality analysis
Every call scored against your checklist, not a 2% sample; coaching cases flagged.
Systems
- Call recordings
- QA forms
MetricShare of calls reviewed
Legal & records
MetricTime per contract review
Contract review
Checks a draft against your checklist and standard terms, marks risky clauses and proposes wording.
Systems
- DMS
- Templates
MetricHours per review
Regulatory search
Questions over laws, regulations and internal policies, answered with the exact clause cited.
Systems
- Regulatory base
- Policies
MetricTime to a cited answer
Deadline control
Renewals, licenses, filings and court dates extracted from documents and tracked.
Systems
- DMS
- Calendar
MetricMissed deadlines
Standard documents
Powers of attorney, letters and standard agreements drafted from templates and case data.
Systems
- Templates
- CRM
- DMS
MetricDrafting time
HR
MetricTime to hire
Résumé screening
Applications checked against the role's criteria and summarized; a recruiter makes every decision.
Systems
- ATS
MetricTime to shortlist
Policy assistant
Staff ask about leave, travel, benefits and procedures and get answers quoted from the policy.
Systems
- Intranet
- HR policies
- Corporate chat
MetricQuestions answered without a ticket
Onboarding
A personal plan for each new hire, reminders for managers, answers to first-week questions.
Systems
- HRIS
- LMS
- Corporate chat
MetricTime to productivity
IT
MetricFirst-line ticket load
One access point to data
Governed, permission-aware access to company data for analytics, instead of ad-hoc exports.
Systems
- Data warehouse
- ERP
- BI
MetricAnalytics backlog
First-line support
Password, access and how-to requests resolved or routed with the diagnosis already done.
Systems
- Service desk
- Directory
- Knowledge base
MetricFirst-line ticket load
Systems documentation
Answers about your own systems from wikis, runbooks and configuration records.
Systems
- Wiki
- CMDB
- Repositories
MetricTime to find an answer
By industry
Use cases by industry
Typical problems we see in each sector, what the platform does about them, and what changes. These are patterns, not case studies; we name clients only with their consent.
Government
Ministries, agencies, municipalities, state enterprises
Typical problems
- Citizen requests bounce between departments
- Directives tracked in spreadsheets and memory
- A regulatory base too large to search
What the platform does
Routes requests to the right agency, tracks directives against deadlines, answers legal questions with citations and gives leadership one console. Runs on-premises or air-gapped with local models.
What changes
Government in detail →Response times become visible, overdue directives surface on their own, and every answer is traceable to its source.
Manufacturing
Plants and industrial groups
Typical problems
- Plan vs. actual known a week late
- Downtime causes logged inconsistently
- Safety violations seen after the fact
What the platform does
Joins production, ERP and sensor exports into one view by line and shift, watches safety on existing cameras and gives process engineers a quoted assistant.
What changes
Deviations are seen in the same shift, downtime has classified causes, and supplier risks are flagged before they stop a line.
Construction & development
Developers, contractors and holdings
Typical problems
- Project knowledge scattered across folders
- Budget overruns found at month-end
- Tender documents rewritten from scratch
What the platform does
One knowledge base over projects, estimates and procedures, schedule and budget control per site, tender drafting from past bids, and a staff assistant in corporate chat.
What changes
Overruns are flagged while they can still be managed, and tender teams start from a draft instead of a blank page.
Financial services
Banks, insurers and lenders
Typical problems
- Manual checks of customer document packages
- High volumes of routine requests and calls
- Strict perimeter and audit requirements
What the platform does
In-app agents that take payments by voice, text or a photo of a document, automated package checks, analysis of requests and calls, all inside the bank's closed perimeter.
What changes
Checks take minutes instead of hours, routine requests resolve without an operator, and every agent action sits in the audit log.
Retail & distribution
Chains, dealers and distributors
Typical problems
- Dealer orders by phone, mail and spreadsheets
- Receivables and stock spread across regions
- Partner programs run by hand
What the platform does
Takes dealer orders automatically, controls receivables and inventory by region, runs loyalty programs for partners and installers, and reports regional sales.
What changes
Orders arrive structured, overdue debt and dead stock are visible weekly, and regional managers stop building reports by hand.
Transport & logistics
Airports, railways and logistics operators
Typical problems
- Passenger flow and occupancy hard to see
- The same questions in many languages
- Maintenance planned from paper records
What the platform does
Video analytics of passenger flow and zone occupancy, multilingual answers for passengers, maintenance planning from asset records, and consolidated analytics for leadership.
What changes
Bottlenecks show up while they form, front-line staff spend less time on repeat questions, and maintenance moves from reactive to planned.
Energy & utilities
Generation, grids, water and district heating
Typical problems
- Call spikes during outages
- Field reports on paper and free text
- Heavy regulatory reporting
What the platform does
Voice agents for outage and billing hotlines, extraction of field reports into structured records, search over asset documentation, and drafts of regulatory reports. Critical infrastructure runs on-premises.
What changes
Hotlines absorb peaks without extra staff, field data becomes usable the same day, and reports start from verified numbers.
Healthcare administration
Hospital groups, clinics and insurers: administration only
Typical problems
- Referral and claims paperwork
- Appointment calls and rescheduling
- Staff questions about procedures
What the platform does
Checks referral and claims documents for completeness, handles appointment requests by phone and chat, and answers staff questions from internal procedures. No clinical decisions.
What changes
Fewer rejected claims for missing data, shorter phone queues, and administrators spend more time on exceptions than on routine.
Telecom
Mobile, fixed-line and internet providers
Typical problems
- Large contact-center volumes
- Churn signals scattered across systems
- Slow triage of network tickets
What the platform does
Multilingual voice and chat agents for support, classification and routing of technical tickets, a weekly churn-risk digest from CRM and billing, and an assistant for tariffs and contracts.
What changes
More requests resolved without an operator, at-risk customers reached earlier, and engineers receive tickets already triaged.
A day with the platform
What a working day looks like
Three roles, one day. Times, names and numbers are illustrative; the pattern is what we build.
Chief executiveExample · illustrative data
Head of a holding or an agency
Reads the morning digest on the phone: one region 12% behind plan, a key supplier contract expiring in nine days, two overdue directives.
Asks: “Why is the North region behind plan this month?” Gets a breakdown by branch and product, with links to the source records.
Opens a one-page brief for the board meeting, assembled from minutes, reports and open issues.
After the meeting, decisions are captured as tasks with owners and deadlines.
Deputies receive their overdue items automatically. Nobody had to chase them.
CFOExample · illustrative data
Finance director of a group of companies
Overnight reconciliations are done: 212 counterparties matched, 9 discrepancies drafted for review.
Receivables: 14 invoices over 60 days, reminders drafted. One write-off above $500,000 waits for approval.
Asks for plan vs. actual operating costs by entity for the quarter, then drills into the largest variance.
Reviews the draft month-end pack; the agent lists three source documents still missing.
Approves flagged payments. Every approval is written to the audit log.
Contact-center leadExample · illustrative data
Head of a multilingual service center
Overnight report: the voice agent handled calls in four languages, 68% resolved without an operator, three escalations.
A spike in calls about a billing change is detected; the answer is added to the knowledge base in minutes.
Quality: all of yesterday's calls scored against the checklist; five flagged for coaching.
Complex cases go to senior operators with a summary already written.
The daily service report goes to leadership without anyone compiling it.
Use case → module → package
Where each use case lives
Each use case runs in a platform module. Packages differ in how many modules and processes they include. The mapping below shows typical fit.
- Included
- Module of choice or add-on
| Use case | Module | Foundation | Enterprise | Sovereign |
|---|---|---|---|---|
| Leadership | ||||
| Plain-language answers about the business | Executive | ✓ | ✓ | ✓ |
| Daily deviation digest | Executive | ✓ | ✓ | ✓ |
| Directive and deadline tracking | Operations | 1 module of choice | ✓ | ✓ |
| Finance & operations | ||||
| Reconciliations and receivables control | Operations | 1 module of choice | ✓ | ✓ |
| Plan vs. actual by shift or branch | Operations | 1 module of choice | ✓ | ✓ |
| Source-document and package checks | Documents | 1 module of choice | ✓ | ✓ |
| Legal, HR and IT | ||||
| Contract review against a checklist | Documents | 1 module of choice | ✓ | ✓ |
| Policy, regulation and procedures assistant | Knowledge | 1 module of choice | ✓ | ✓ |
| First-line internal support | Knowledge | 1 module of choice | ✓ | ✓ |
| Contact center and sites | ||||
| Multilingual voice agents | Voice | — | Add-on | ✓ |
| Safety, flow and occupancy video analytics | Vision | — | Add-on | ✓ |
| Scope | ||||
| Business functions | One | 3–5 | Organization-wide | |
| Processes | Up to 3 | Up to 15 | Unlimited | |
Foundation includes the Executive module plus one module of your choice; Enterprise includes all core modules, with Voice and Vision as add-ons; Sovereign includes every module. Exact scope is fixed in the proposal. See packages and the full comparison.
Choosing the first process
What makes a good pilot
The signs we look for during the diagnostic. The more of them a process has, the faster the pilot pays back.
01 · Frequency
It happens every day
Dozens or hundreds of similar operations a week.
02 · Data
The data already exists
Even if it's spread across systems and spreadsheets.
03 · Measurability
It's clear how to count
Hours, days, errors, money: something you can compare before and after.
04 · Ownership
Someone owns it
A person this process frustrates, who wants it changed.
05 · Risk
Mistakes are recoverable
The first process shouldn't be one where a single error is expensive.
06 · Visibility
The result shows
Leadership should be able to see the pilot's impact and decide on scaling.
Regional notes
Europe, the Americas and Asia-Pacific
The use cases are the same everywhere; languages, data-residency rules and the usual deployment differ. We design every deployment to support compliance with local rules; your legal team keeps the final word.
| Region | Working languages | Rules we design for | Typical deployment |
|---|---|---|---|
| Europe | English, German, French, Spanish, Portuguese and others | GDPR, UK GDPR, EU AI Act | BlackGust Cloud in an EU region, private cloud or on-premises |
| United States & Canada | English, Spanish, French | CCPA/CPRA and other state privacy laws, HIPAA for health administration, PIPEDA | In-region cloud, your private cloud, on-premises for the public sector |
| Latin America | Spanish, Portuguese | LGPD in Brazil, national data-protection laws elsewhere | In-region cloud, private cloud or on-premises |
| China | Chinese | PIPL, Data Security Law, rules for generative AI services | On-premises or private cloud inside mainland China, local models |
| Japan | Japanese | APPI | In-region cloud, private cloud or on-premises |
| Korea | Korean | PIPA | In-region cloud, private cloud or on-premises |
| Australia | English | Privacy Act 1988 and the Australian Privacy Principles | In-region cloud, private cloud or on-premises |
More on data residency and deployment options on the security page.
FAQ
Questions about use cases
Which use case should we start with?
The one that happens every day, has data, can be measured and has an owner who wants it changed. The diagnostic scores 3–5 candidates on exactly these points and recommends one; see how we choose.
Our problem isn't on this page. Can you still solve it?
Probably. The library shows what we see most often, not everything we do. If the work involves reading, checking, routing or reporting on information in your systems, it is usually a fit. Describe it in a request and an engineer will answer.
Do we need clean, centralized data before we start?
No. Most clients start with data spread across an ERP, spreadsheets and mailboxes. The platform connects to where it lives. If records are missing entirely, we say so in the diagnostic and recommend fixing record-keeping first.
Which systems can you connect?
ERP and accounting systems such as SAP, Oracle and Microsoft Dynamics, CRMs such as Salesforce, document stores, mail, telephony, databases, spreadsheets and cameras. If a system has an API, a database replica or a file export, we can usually connect it.
How quickly does a use case pay back?
We don't promise a number before the diagnostic. It produces an ROI estimate with a conservative and an expected case for your process, and the pilot measures the real result against a baseline.
Will the AI make decisions on its own?
Only where you allow it. By default agents prepare and propose; a person approves actions above $500,000 or when the agent is not confident. Every step is logged.
Is our data used to train AI models?
No. Cloud models are used through enterprise APIs whose terms exclude training on customer data, and local models run inside your infrastructure. See security.
Do you work with government and regulated industries?
Yes. Public bodies, banks, energy companies and healthcare administrators usually choose on-premises or air-gapped deployment with local models. See AI for government.
Which languages do the agents work in?
Text agents work in all major languages. Voice agents support English, Spanish, Portuguese, German, French, Chinese, Japanese and Korean, among others.
Related
Keep reading
Platform
Modules, agents and the data model
What runs underneath every use case on this page.
Open →ApproachFrom diagnostic to scale
How forward-deployed engineers take a use case to daily use.
Open →GovernmentAI for the public sector
Sovereign, on-premises deployment for ministries and agencies.
Open →PricingPackages and scope
Foundation, Enterprise and Sovereign compared.
Open →Don't see your problem here?
Describe it in your request. An engineer will tell you whether it's solvable and where to start.