Verity for data companies

You deliver data other people trust. Verity helps you prove it is right, on time and safely handled.

A data services firm sells accuracy, throughput and care with someone else's data. Guidelines change mid-project, quality is checked in samples and payment depends on a client accepting the batch. Verity gives every project one record of instructions, people, quality, delivery and access.

Verity in one line

One system that follows each project from client guidelines to annotator work, quality checks, delivery and acceptance, and shows who is slipping and what is still unbilled.

See your data company in Verity See the 14 problems we solve

Free: we set up a working demo of your own business, on your own data, within 7 days. You pay only if your team decides to use it.

Built for a firm with managers, team leads and a large floor. Records come from the project work itself.

Verity / Firm / Today

All projects

Live

Projects live

9

4 clients

At risk of late

3

projects

Review coverage

11%

target 15%

Accepted, not invoiced

₹9.6 L

6 batches

Cost per accepted unit against price

  • Project A62%
  • Project B88%
  • Project C106%
  • Project D74%

Project C costs 106 percent of its price per accepted unit after rework. Review coverage is below target on 2 projects.

Review
  • Tool seats for Client 314 seats unused since project endedRelease
  • Deletion of Client 2 dataDue 20 Oct, 3 machinesAssign

Illustrative screen with sample figures. Yours is set up with your own clients, projects, teams and queues.

How it runs today

Most data firms track volume well and quality loosely.

The work is counted in units: images, documents, minutes or records. Volumes are in a tool, and instructions are in a document that changes. Quality is checked by sampling, and feedback arrives by email.

The people are many and the turnover is high. One 2026 report on annotators described pay of ₹10,000 to ₹15,000 a month, strict targets and software that monitors speed and accuracy, so performance data matters, and so does how it is used.

The commercial pressure is real. A 2026 industry page says high volume commitments compress quality review cycles, and that hundreds of vendors compete with few shared quality frameworks.

What most data companies run on

  • Annotation or work tools from each client
  • Instruction documents and chat groups
  • Spreadsheets for output and quality
  • Attendance and payroll sheets
  • Email for client feedback

Each tool shows one slice. None of them shows whether a project makes money, which team is slipping on quality or which client data is still on a machine it should not be.

14 problems, one by one

Where a data company loses margin, quality and trust.

These are the situations data firm owners and delivery heads describe. For each one: what it looks like, what it costs, and what Verity does about it.

  1. 01Projects

    Guidelines change and the team works from the old version

    What it looks like

    A client revises the labelling rules on day six. The update goes to a chat group. Half the team keeps working to the old version for a day.

    What it costs you

    A batch with inconsistent labels, rework at your cost and a quality complaint that is hard to explain.

    How Verity fixes it

    Each project holds its guidelines as versions with the date they took effect. A change creates a task for team leads to brief their teams and confirm, and work is tagged to the version it followed.

    See the screen
  2. 02Quality

    Throughput targets squeeze quality checks

    What it looks like

    The client wants 40,000 items by Friday. Review is the first thing cut so the target is met.

    What it costs you

    High-volume commitments compress review cycles, and the errors appear when the client audits. The cost of rework and lost trust is higher than the review time saved.

    How Verity fixes it

    Each batch shows items done, items reviewed and error rate. Verity warns when review coverage falls below the level you set before a batch is sent.

    See the screen
  3. 03People

    Annotator performance is measured, but not fairly or usefully

    What it looks like

    Speed is tracked and accuracy is sampled. A new hire is compared with a veteran and a slow week is read as a poor worker.

    What it costs you

    Good people leave. In one 2026 report annotators described strict targets and monitoring by software, with pay of ₹10,000 to ₹15,000 a month.

    How Verity fixes it

    Output and accuracy are shown per person by task type and by week, with experience and project difficulty alongside. Leads see trends and coaching needs, not only a red cell.

    See the screen
  4. 04Projects

    Work queues and deadlines are juggled by hand

    What it looks like

    Five projects share one floor. A team lead moves people between queues by feel. A deadline is missed because a queue was drained too slowly.

    What it costs you

    Late deliveries, overtime and idle people on a different project.

    How Verity fixes it

    Each project has a queue of work, a deadline and an allocation of people. Verity shows queue size, daily output and the date the batch will finish at the current pace.

    See the screen
  5. 05Quality

    Error types are not categorised, so the same mistakes repeat

    What it looks like

    Reviewers mark items wrong. The reasons are in comments and nobody counts them.

    What it costs you

    The same error is made again by many people, and coaching is generic.

    How Verity fixes it

    Reviewers choose an error type when they mark an item. Verity shows common errors by project, person and guideline section, so fixes aim at the cause.

    See the screen
  6. 06Money

    Projects are priced per unit and nobody knows the real unit cost

    What it looks like

    A project is sold at ₹4 per item. Rework, review, tool time and idle hours are not added up.

    What it costs you

    Projects that look busy and lose money. Throughput pressure and many competing vendors make price cuts tempting.

    How Verity fixes it

    Each project tracks units delivered, hours spent by role, rework and tool cost. Verity shows cost per accepted unit against the price.

    See the screen
  7. 07Delivery

    Client acceptance is slow and disputed

    What it looks like

    A batch is delivered. The client takes three weeks to review and rejects 6 percent with a spreadsheet of comments. The invoice waits.

    What it costs you

    Cash held up by acceptance and disputes with no record of what was delivered and when.

    How Verity fixes it

    Each delivery records items, the version, the date and the client's acceptance or rejection with reasons. Rejected items become rework tasks and invoices follow accepted units.

    See the screen
  8. 08Security

    Client data spreads onto machines and chats

    What it looks like

    A client shares sensitive documents. Annotators view them on shared machines and a team lead forwards a sample to ask a question.

    What it costs you

    A leak ends a client relationship and may breach the contract. Data privacy concerns in sectors like banking and healthcare can require on-shore delivery.

    How Verity fixes it

    Each project records what data it holds, who has access, where it may be processed and when it must be deleted. Access is removed when a person leaves a project, and a log shows who opened what.

    See the screen
  9. 09People

    Domain experts and language reviewers are scarce

    What it looks like

    A medical imaging project needs reviewers with a clinical background. Only two are available and both are booked on another project.

    What it costs you

    Delays or reviews done by people who lack the skill, and quality that suffers for the sensitive part of the work.

    How Verity fixes it

    People have skills, languages and certifications recorded. Verity shows who can review a type of work and their load, so scarce reviewers are planned.

    See the screen
  10. 10Projects

    Tools and licences are tracked per client

    What it looks like

    Each client asks for a different annotation tool. Licences and seats are bought, and some stay unused after a project ends.

    What it costs you

    Money spent on seats nobody uses and delays when a person has no access on day one.

    How Verity fixes it

    Each tool and its seats are records linked to projects and people. Verity lists unused seats and projects that need access.

    See the screen
  11. 11Delivery

    Recurring data deliveries and refresh cycles are scheduled in calendars

    What it looks like

    A client buys a dataset refreshed monthly. A delivery date slips because the source update was late and nobody told the client.

    What it costs you

    A client who counts missed refreshes at renewal and credits that are argued.

    How Verity fixes it

    Each recurring delivery has a schedule, an owner and checks before it ships. A late source creates a task, and a renewal list shows what each client has received.

    See the screen
  12. 12Money

    Invoices follow memory, not accepted work

    What it looks like

    Invoices are raised from a spreadsheet of units. Some include items later rejected, some leave out accepted items, and a foreign client pays late.

    What it costs you

    Credit notes, delayed cash and an accounts team that reconciles by hand.

    How Verity fixes it

    Billable units come from accepted deliveries. Verity lists work accepted and not invoiced, and invoices overdue by client, with a task for the account owner. Billing stays in your accounts software.

    See the screen
  13. 13Compliance

    Personal data duties are written, not tracked

    What it looks like

    Contracts say the firm will delete data at the end and report any breach. Nobody can show that it did.

    What it costs you

    A client audit that fails, or a dispute after an incident. India's data protection rules were notified in November 2025, with most duties applying over the next 12 to 18 months.

    How Verity fixes it

    Each project has deletion dates, breach contacts and the steps to follow. Verity lists projects whose data is due for deletion and records that it was done.

    See the screen
  14. 14People

    Hiring and training come in waves

    What it looks like

    A large new project needs 80 people in two weeks. Hiring, training and quality ramp-up are tracked in separate sheets.

    What it costs you

    New people on live work before they are ready, and early errors that hurt a new client's first impression.

    How Verity fixes it

    Each batch of new people has a training plan, assessments and a probation period. A person starts live work only after the assessment, and early quality is tracked separately.

    See the screen

The numbers behind it

The numbers that make these problems worth fixing.

From published sources, with the source named under each figure. Your own numbers will differ, and showing you yours is the first thing Verity does.

  • About 35%

    India's share of the global data annotation market, according to one industry page, which also projects the global market at USD 5.3 billion or more by 2028.

    HanDigital, 2026

  • ₹10,000 to ₹15,000

    the monthly pay reported for annotators at one firm, with strict targets and software monitoring of speed and accuracy.

    ThePrint, 2026

  • 65% to 70%

    of revenue in India's data annotation market contributed by managed service providers, according to a Nasscom analysis.

    Nasscom, 2025

  • About 70,000

    professionals engaged full and part time in data annotation in India in 2025, according to Nasscom's report.

    Nasscom, 2025

Industry size figures vary widely between sources and none is used. The first figure is a commercial page, the pay figures are from one reported firm, and the workforce figure could not be checked against the original page. They describe the sector, not what your firm will see.

What you will actually see

The screens you would use every day.

Six views, each for one job. Clients, projects, teams and queues are set up in your own terms.

Screen 1 of 6

The delivery head's daily view

Queues, deadlines, quality, people on each project and work accepted but not yet invoiced on one screen, with the items that need a decision listed below.

Fixes

  • 06 Projects are priced per unit and nobody knows the real unit cost
  • 10 Tools and licences are tracked per client

Verity / Firm / Today

All projects

Projects live

9

4 clients

At risk of late

3

projects

Review coverage

11%

target 15%

Accepted, not invoiced

₹9.6 L

6 batches

Cost per accepted unit against price

  • Project A62%
  • Project B88%
  • Project C106%
  • Project D74%

Project C costs 106 percent of its price per accepted unit after rework. Review coverage is below target on 2 projects.

Review
  • Tool seats for Client 314 seats unused since project endedRelease
  • Deletion of Client 2 dataDue 20 Oct, 3 machinesAssign

Screen 2 of 6

Projects, guidelines and queues

Each project with its guideline version, queue, deadline and allocation. A change in guidelines creates a task for every team lead to brief and confirm.

Fixes

  • 01 Guidelines change and the team works from the old version
  • 04 Work queues and deadlines are juggled by hand

Verity / Projects

Live work

Briefing1

  • P-41 Medical images, Client 4Guidelines v3 issued, 2 of 5 leads confirmedConfirm

In production2

  • P-38 Retail receipts, Client 128,400 of 40,000, due FriAt pace
  • P-36 Call transcripts, Client 2Queue 12,000, finishes 3 days lateLate

Delivered1

  • P-33 Street scenes, Client 3Awaiting acceptance 9 daysChase

Screen 3 of 6

Review coverage and error types

Review coverage, error rate and the types of error by project and by guideline section, so coaching and fixes aim at causes.

Fixes

  • 02 Throughput targets squeeze quality checks
  • 05 Error types are not categorised, so the same mistakes repeat

Verity / Quality

This week

ProjectItemsReviewedError rateTop error
P-38 Receipts28,4009%3.8%Wrong field
P-36 Transcripts18,90014%2.1%Speaker label
P-41 Medical3,20025%5.4%Boundary
P-33 Street scenes22,00012%1.6%Occlusion

P-38 review coverage is 9 percent against a 15 percent target.

Add reviewers

Screen 4 of 6

People, skills and onboarding

Output and accuracy by person and task with experience alongside, scarce reviewers by skill and the training and assessment status of new batches.

Fixes

  • 03 Annotator performance is measured, but not fairly or usefully
  • 09 Domain experts and language reviewers are scarce
  • 14 Hiring and training come in waves

Verity / People

Teams and training

PersonTaskOutputAccuracyNote
Anita (new, week 3)Receipts380 a day94%Coach
Ravi (2 years)Receipts610 a day97%Strong
Dr. Shah (reviewer)Medical90 a day99%Booked 95%
ImranTranscripts230 a day96%On track
  • Batch 14, 24 people, assessment passed22 of 24
  • Two people to repeat the assessmentFriday

Screen 5 of 6

Delivery, acceptance and invoicing

Batches delivered with the client's acceptance or rejection, rework tasks, recurring deliveries and what is accepted but not yet invoiced.

Fixes

  • 07 Client acceptance is slow and disputed
  • 11 Recurring data deliveries and refresh cycles are scheduled in calendars
  • 12 Invoices follow memory, not accepted work

Verity / Delivery

Batches and invoices

BatchClientDeliveredAcceptedStatus
P-33 B4Client 39 days agoPendingChase
P-36 B2Client 221 days ago94%6% rework
P-38 B1Client 13 days agoPendingWaiting
Monthly dataset, OctClient 5Due 28 OctNot yetSource late

₹9.6 lakh is accepted but not invoiced across 6 batches.

Raise invoices

Screen 6 of 6

Data access, deletion and evidence

Which projects hold which client data, who has access, where it may be processed and when it must be deleted, with a record that deletion happened.

Fixes

  • 08 Client data spreads onto machines and chats
  • 13 Personal data duties are written, not tracked

Verity / Security

Client data

ProjectDataPeople with accessDelete byStatus
P-41 MedicalMedical images9Project endOn-shore only
P-36 TranscriptsCall audio2231 OctDue
P-29 Receipts (ended)Scanned receipts320 OctOverdue

Three people still have access to project P-29, which ended on 5 Oct.

Remove

Illustrative screens with sample data. Layouts, fields, names and terms are set up around how your data company works.

An ordinary day

The same Tuesday, before and after.

The floor works the same way. What changes is that every project leaves a record that chases for you.

WhenTodayWith Verity
9:30 TodayThe delivery head asks team leads for queue sizes on a call. With VerityThe Today screen shows queues, pace and which projects will finish late.
11:00 TodayA guideline change reaches half the team. With VerityEvery lead had to confirm the briefing and the work is tagged with the version followed.
1:00 TodayA batch goes out with thin review. With VerityA warning stopped it at 9 percent coverage and reviewers were added.
3:30 TodayA client rejects 6 percent of a batch and nobody knows which items. With VerityRejected items are rework tasks with reasons and owners.
5:30 TodayAccess for a finished project is left open. With VerityAccess was removed with the project close, and the log shows when.
Month end TodayFinance builds invoices from a unit sheet and a client dispute follows. With VerityThe accepted-not-invoiced list drives the invoice.

How we work with you

We set it up with you, in four weeks.

You do not get a login and a manual. We learn how your projects actually run, build Verity around them, bring your data in, then stay on to keep it working.

  1. Week 1

    Discovery and mapping

    What we do
    Spend a day with your delivery head, a team lead and the person who invoices. Follow a project from client brief to acceptance, and map your tools, quality checks and security controls.
    What we need from you
    A few hours from you and your delivery head. Two recent projects with guidelines, output and quality sheets and invoices.
    What you have at the end
    A written map of how projects, people, quality and delivery run today and the problems to fix first.

  2. Weeks 2 to 3

    Configuration

    What we do
    Set up projects and guideline versions, queues and allocations, quality review rules and error types, people with skills, delivery and acceptance stages, data access and deletion rules and approvals.
    What we need from you
    Check the review rules, error types and deletion rules with your delivery head and a team lead.
    What you have at the end
    A working Verity with your real project types, teams and rules.

  3. Weeks 3 to 4

    Migration and training

    What we do
    Bring in clients, live projects, people and skills, open batches and data access lists. Train managers, leads and the invoicing owner on live work.
    What we need from you
    Your list of live projects and people, in whatever form they exist. One hour per role.
    What you have at the end
    Your firm inside Verity and a team that has used it on a live project.

  4. After go-live

    Operations partnership

    What we do
    Stay on after launch. Review the first weeks with you, tune lists and thresholds and handle new clients, locations and shifts.
    What we need from you
    A short monthly review of what the numbers say.
    What you have at the end
    A system that keeps matching how the firm runs, with someone responsible for it.

Bring these on day one

  • List of live projects with clients and deadlines
  • Current guidelines and versions
  • People list with skills and tools
  • Quality sheets and error categories you use today
  • Your data security and deletion commitments by client

Your time

About 8 to 11 hours of owner time across the four weeks, mostly in week 1. Your delivery head and team leads need a few more.

After go-live

Nobody is hired to run Verity. Output, quality and delivery are recorded as part of the work you already do, and the reports come from those records.

Verity AI

Ask your firm a question.

Verity AI reads your projects, quality, people, deliveries and security records. Ask in plain language, then turn the answer into tasks for the right person.

  • Your records only Answers come from your own data companies data, not from generic knowledge.
  • Right person, right view It only shows what the person asking is allowed to see.
  • Turns into action An answer can become a task assigned to a named person.

Verity / Ask

Which projects will miss their deadline at the current pace?

Three. P-36 will finish three days late and P-41 two days late.

  • P-363 days late
  • P-412 days late
  • P-381 day late

Create allocation tasks for the team leads

Verity / Ask

Where is review coverage below target, and what are the main errors?

P-38 at 9 percent against 15. The main error is wrong field, 41 percent of mistakes.

  • P-38 coverage9%
  • Wrong field41% of errors

Add reviewers and brief the team

Verity / Ask

Which closed projects still have people with access to client data?

One. P-29 ended on 5 Oct and three people still have access.

  • P-293 people

Create access removal tasks

Sample questions and answers with illustrative figures.

Without chasing

The follow-ups nobody has time to make.

Set once, then they run from your records. People still do the work. Verity makes sure the right person knows to.

When

A project guideline changes

  • Team leads asked to brief and confirm
  • Work tagged to the version followed
  • Unconfirmed leads listed after 24 hours

When

Review coverage on a batch is below target before delivery

  • Delivery held
  • Task to add reviewers
  • Release recorded with the final coverage

When

A delivered batch is not accepted within 10 days

  • Added to the chase list
  • Task to the account owner
  • Outcome recorded

When

A project closes

  • Access removal tasks created
  • Deletion date set
  • Completion recorded with who did it

When

A person's accuracy falls below the set level for a week

  • Flagged to the team lead
  • Coaching task created
  • Result recorded

When

A tool seat is unused for 30 days

  • Listed for release
  • Task to the tool owner
  • Saving recorded

Rules and paperwork

The rules a data company has to keep up with.

Verity does not give legal advice and does not file anything for you. It keeps the records these rules and your clients' contracts depend on in order.

  • Personal data you process for clients

    What is expected

    India's data protection rules were notified in November 2025, and most duties start to apply over the following 12 to 18 months. The client who decides why data is processed remains responsible for what a processor does on its behalf, so clients will ask what you do.

    How Verity helps

    Each project records what data it holds, who has access, where it is processed and when it is deleted.

    AZB Partners

  • Client contracts and confidentiality

    What is expected

    Each contract sets the guideline version, quality level, security terms and deletion duties. A breach can mean credits or termination.

    How Verity helps

    Terms sit on the project and are compared with delivery and access records.

  • On-shore or in-country delivery

    What is expected

    Data privacy concerns in sectors such as banking and healthcare may require work to be done in specific locations.

    How Verity helps

    Each project records where data may be processed, and people and machines are assigned only to the allowed places.

    Nasscom

  • Working hours and staff records

    What is expected

    A 2026 report described long screen hours, strict targets and few paid leaves at one firm. State shop and establishment rules and wage laws set limits, and records are what show you comply.

    How Verity helps

    Attendance, shifts and leave are recorded per person.

    ThePrint

Rules and client requirements differ by sector and contract, and they change. Confirm current requirements with your lawyer and your clients.

Who sees what

Everyone sees the same records, through their own window.

Annotators should not see client invoices. Finance should not see raw client data. Each role opens on what it needs.

  • Founder or managing director

    Which projects and clients earn, and where are the risks?

    Cost per accepted unit, work not invoiced, deadlines at risk, security and deletion status.

  • Delivery head

    What will miss its date and why?

    Queues and pace, allocation, review coverage, error types and rework.

  • Team lead

    How is my team doing today?

    Their queue, output and accuracy, guideline briefings to confirm and coaching tasks.

  • Quality reviewer

    What do I check next?

    Items to review, error types to choose, feedback to give and reviews due by project.

  • Finance owner

    What is accepted and what is owed?

    Accepted not invoiced, invoices overdue, tool costs and project margins.

Straight answers

What changes, and what stays the same.

Keep using

  • Your annotation and labelling tools and each client's tools
  • Your attendance device and payroll process
  • Your accounts software and accountant
  • Your secure file transfer and storage

Verity does not

  • Label, annotate or store raw client data
  • Run your annotation tools or set client guidelines
  • Send messages to clients or staff on its own
  • Give legal or privacy advice

We bring your clients, projects, people, queues and access lists across from the sheets, chats and tools you use now, in stages, so work keeps flowing while it happens.

  • Excel
  • Google Sheets
  • Legacy ERP
  • CRM
  • One operating environment

Questions

Questions data company owners ask.

Does Verity replace my annotation tool?

No. Your annotation tools stay as they are. We agree during setup how output and quality results reach Verity, so you see projects, people, quality and delivery in one place.

Will team leads and reviewers use it?

It works only if recording takes seconds, so we set it up around the sheets and checks they already use. They get a clear view of their queue and their team's quality.

Can Verity measure my annotators?

It shows output and accuracy per person with experience and task type alongside. How you use it is your call, and the system supports coaching as well as targets.

Does it hold client data?

No. It records which data each project holds, who has access and when it must be deleted. The data itself stays in your secure systems.

Can it help with invoicing?

Yes. It lists work accepted and not invoiced and invoices overdue. Billing stays in your accounts software.

I run two sites and a remote team. Can I see all of it?

Yes. Each site is a location in one system, and projects, people and quality can be compared side by side.

Can it send reports to my clients?

Not by itself. It tracks delivery, acceptance and rework, and your team sends the reports and records the date.

Who can see what?

Only the people you allow. Access is set by role and by project, and every change is recorded with who made it and when.

What does it cost?

It depends on the number of users and what you want set up. Message us and we will give you a clear number after a short conversation about how your firm runs.

Where the figures come from

Sources

  1. HanDigital on the annotation industry, 6 Apr 2026
  2. ThePrint, 1 Sep 2026
  3. Nasscom
  4. AZB Partners on the DPDP framework

Last checked 16 October 2026. Published figures vary by region and method, so treat them as indicative, not as a promise of your results.

See which projects really make money.

Tell us how your firm runs and we will show you the thin review coverage, late projects and unbilled work it would find, using your own numbers.

Get your free demo Email us

Free: we set up a working demo of your own business, on your own data, within 7 days. You pay only if your team decides to use it.