From Research to Finished Work: Turning Opis Intelligence into Something a Client Can Use

Start with the research asset, not the final document

When an Opis run completes, the useful thing underneath the visible report is a structured Knowledge Object.

It can contain the analysis itself alongside sources, confidence, assumptions, contradictions, ambiguities, gaps, resolved gaps, scoring information where relevant, timelines, source relevance and refinement history.

That structure gives downstream features something more useful than a block of finished prose.

The same research can be presented differently without having to rediscover its meaning each time.

That is the foundation for client-ready production.

Step 1 — White-glove the output for the client

An agency may research ten clients in the same week.

Those clients should not all receive documents that look like they belong to the agency's software vendor.

Brand Presets allow Opis to retain reusable client/company presentation settings: names, colours, logos, typography, chart palettes, report styling and preferred report languages.

The aim is operational consistency.

Configure the brand once, then reuse it across future work rather than manually restyling every report.

The design tools can also assist with extracting a starting palette from a client website or visual reference. That should still be treated as a starting point: a human should verify that the colours, typography and identity actually represent the brand before publishing client work.

Once a report is generated, sections can be edited and styled for the specific deliverable.

Research discipline and presentation discipline are separate jobs. Opis tries to give you control over both.

Step 2 — Choose the format the recipient will actually use

Not everybody wants an interactive dashboard.

A client board may want a PDF. An analyst may want CSV. A colleague may want a DOCX they can continue editing. An automation may prefer structured JSON plus a rendered file.

Opis's export layer supports multiple formats, including PDF, DOCX, HTML, Markdown, text and CSV where the content permits it.

This is not glamorous functionality, but it matters enormously in professional work.

The “best” report format is the one that enters the recipient's existing workflow with the least friction.

Do not send an interactive link merely because it looks impressive if the procurement team needs an attached PDF for its records.

Do not flatten structured information into a PDF if an automation needs machine-readable fields.

Delivery should follow the job.

Step 3 — Transcreate without rebuilding the analysis

International work creates a difficult problem.

The English report may be correct, yet a literal translation can sound awkward, culturally wrong or too informal for the target audience.

At the same time, allowing a language model to freely rewrite the entire analytical object creates another risk: numbers, scores and structural relationships can drift during translation.

Opis separates the canonical analysis from its language presentation.

Its transcreation system can build locale-oriented style guidance and apply that guidance to the human-readable language while preserving the underlying analytical structure.

The current output/transcreation registry supports 64 locales. That is different from saying every part of the website UI is already translated into 64 languages; it is the research/content output capability.

That distinction is worth keeping clear.

For an international agency, the practical workflow can become:

Research once → verify once → adapt the communication for each target market.

You should still have important client-facing work reviewed by someone who understands the destination language and market. Transcreation reduces the mechanical work; it does not make cultural judgement obsolete.

Step 4 — Turn intelligence into content

Suppose your research has established:

• a recurring customer frustration;
• a demonstrable competitor weakness;
• an attractive product feature;
• evidence supporting the proposition;
• the language used by the intended customer.

That is exactly the information a good content team would want before writing anything.

Content Studio gives the research another destination.

Instead of beginning from a blank prompt, it can begin from completed Opis intelligence or a collected profile. It supports registered formats for LinkedIn, X, Facebook, Instagram and newsletters, and allows the generated pieces to be edited rather than treating the first draft as sacred.

This makes the content workflow:

Research → proposition → platform format → edit → validate → transcreate if required → deliver.

That is much more defensible than generating twenty generic social posts and hoping one sounds relevant.

Step 5 — Deliver without turning Opis into a walled garden

The work often needs to leave Opis.

Native delivery paths support destinations such as email, Telegram, Slack, Discord and generic webhooks. Webhook destinations can be labelled for services such as Zapier, Make and n8n, allowing Opis output to enter wider automation flows.

There is also an external API-key mechanism for inbound automation. An outside system can authenticate using an App Hub-issued key and trigger supported Opis research through the same execution path used by the product, so entitlement and quota rules still apply.

Together, those two directions matter:

External system → Opis research → structured intelligence → external destination.

That is the foundation for more advanced operational use cases.

Where Claude Code fits—and where it does not

This deserves very precise language.

Claude Code is not inside Opis.

Opis does not sell Claude Code, embed Anthropic's coding agent as its own feature or give every user an invisible Claude worker.

For users who already choose to use Claude Code and its browser capabilities, Opis can provide an external automation surface: API access, webhook destinations and an integration starter pack can give the external agent something useful to operate.

For example, a user could ask their own Claude Code environment to help configure a Zapier workflow that takes an Opis output and routes selected fields into another system.

Likewise, external scheduling belongs to the user's automation environment. A user may choose to schedule a task through their own Claude Code/cron setup and have that automation trigger Opis. That is different from claiming Opis itself contains a native overnight Claude scheduler.

The distinction protects both products and makes the architecture easier to understand.

Opis provides the intelligence capability.

External automation decides when and where to use it.

An outcome-led example: the morning sales brief

Imagine a small agency or sales team using HubSpot.

A new qualified lead enters the CRM in the evening.

An external automation detects the event and triggers an appropriate Opis research run using the information already available about the company.

Opis investigates the defined research question and produces a structured result with evidence, uncertainty and sources.

The automation takes the useful fields and returns them to the team's working environment—perhaps a CRM record, Slack channel or another destination configured through Zapier or a webhook.

The salesperson arrives the next morning with research waiting for them.

Notice what is being sold in that story.

Not “AI agents.”

Not “webhooks.”

Not “an API.”

The outcome is:

The salesperson starts the day knowing more about the opportunity without spending the first hour researching it.

That is the level at which automation becomes commercially meaningful.

Another outcome: the multilingual client campaign

Now consider a content agency working for a product client.

The team runs competitor and customer-friction research, refines weak assumptions through ICC and combines the strongest sections into a collected profile.

That profile becomes context for Content Studio.

The team creates platform-specific content, edits the final pieces, applies the client's saved brand direction and transcreates selected output for a second market.

The finished assets are exported or delivered to the appropriate downstream workflow.

Again, the important part is not that Opis has a report button and a social button.

The important part is that the same researched intelligence survives through both.

Your delivery exercise

Take one research result you already trust.

Then deliberately move it through the production chain:

1. Reopen it from the Centre Library.
2. Confirm its important sources and uncertainty markers.
3. Apply or create the relevant Brand Preset.
4. Edit one section for the actual recipient rather than accepting the default copy.
5. Export it in the format the recipient would genuinely use.
6. Send the same research into Content Studio and create one appropriate piece—not ten generic ones.
7. If a second language matters, transcreate that piece and review it.
8. Decide whether it should be delivered manually or through an existing integration.

This exercise teaches something that a feature tour cannot.

It shows you where Opis fits in your real work.

The point of the system

Research software often optimises the moment when an answer appears on screen.

Professional work has a longer lifecycle.

Someone must question the evidence.

Someone must decide what matters.

Someone must adapt the result to a client.

Someone may need to translate it, turn it into content, archive it or move it into another system.

Opis becomes more useful when those activities are treated as one continuous information journey rather than a collection of disconnected AI tricks.

The research engine is the beginning.

The outcome is the product.