Fernanda Oi is a UX Research Manager at Venturus. Oi's work includes UX research and product discovery: helping teams understand users and make better-informed product and service decisions before development.
Fernanda Oi is most useful for UX research and product discovery before development.
Best for
B2B SaaS brands improving product discovery with custom UX research and design.
Not for
Broad consumer awareness campaigns needing large reach and measured audience delivery.
Key takeaways
Test AI research tools on a real question before trusting them.
Discovery aims to find problems before development, scoped to each challenge, not a fixed recipe.
Build the research function, then introduce it to the whole company.
Check first: Confirm paid terms, available formats, audience delivery, and recent post results before contacting.
Main ideas
Test AI research tools on a real question before trusting them · Discovery aims to find problems before development, with work scoped to each challenge rather than a fixed recipe · Build the research function, then introduce it to the whole company
Fernanda Oi is a UX Research Manager at Venturus whose posts show two sides of research work: shaping product discovery before development and building a research team inside an organization.
In a post about a co-authored article, they describe Discovery as tailored work using UX Research and UX Design to help teams make better-informed product and service decisions.
A separate post shows them testing AI research tools with the same prompt, checking source quality, date limits, depth, and whether follow-up reports add distinct insights.
These practical examples make their work worth following for people interested in how research is planned, evaluated, and organized.
The clearest strength in the posts is making research quality concrete.
Rather than simply recommending an AI tool, Oi describes a repeatable comparison: use the same prompt, identify where results fall short of the intended task, test several related reports for duplication, then look for patterns across them.
They also frame product Discovery as tailored to the challenge rather than a fixed recipe.
Together, these examples show a practical interest in both the quality of research outputs and the conditions that help teams use research before development.
The posts also acknowledge limits: Gemini sometimes stalled and produced overly long text.
Fernanda Oi's posts focus on doing research well under real constraints: testing research tools against an actual question before trusting them, scoping discovery work to surface problems before development and building a research team the wider company can see.
Read more
The through-line is that depth and speed are a design problem to solve, not a tradeoff to accept, and that credit for the work is shared rather than claimed alone.
Test AI research tools on a real question before trusting them
Fernanda Oi argues that teams should choose an AI deep-research tool based on evidence, not habit.
Read more
The test uses the team's own question, with three deciding criteria: whether sources fall inside the requested date range, whether they are primary or academic rather than sites that merely cite research, and whether the final text draws connections across findings instead of listing them.
The conclusion: ChatGPT is not automatically the best or only option.
The preference is conditional; the same post notes that the tool that performed better still freezes at times and returns over-long texts.
The argument sits within a wider priority: research efficiency means going deeper without sacrificing time or quality, since more information usually takes more time and deadlines are tight.
They ran the same prompt — behavioural data on a defined user group, single people with children, drawing on scientific and practitioner sources in Portuguese and English from 2020 to 2025, correlating neurobiology, behaviour, psychology and cognition — through OpenAI's Deep Research and Gemini's Deep Research.
OpenAI returned a 1987 article outside the window, leaned on sites that only cited studies and produced shallow text with no correlations.
A colleague, Victor Scarpa, suggested trying Gemini, which returned an editable research plan, stayed closer to the prompt, used academic sources and listed relevant sources it had not used.
Fernanda Oi then extended the test with three reports on coparenting from different angles and asked for a cross-report analysis of patterns and outliers.
The result was a new reading that referred back to the earlier reports without collapsing into a summary.
Discovery aims to find problems before development, with work scoped to each challenge rather than a fixed recipe
Fernanda Oi argues that discovery work makes a product or service more likely to land by surfacing issues before development rather than after.
Read more
The method applies UX research and UX design techniques to each engagement, matching a specific challenge to a solution quickly.
The work is customised rather than run from a standard playbook.
For buyers, the deliverable is assertiveness and knowledge about users, not a packaged sequence of steps.
They announced an article co-authored with Antonio Salgado describing Discovery as one of the services in Venturus's design portfolio and pointing readers to it for more detail. The article itself was not part of the material reviewed here, so this reflects the announcement's framing of the service rather than the article's contents.
Build the research function, then introduce it to the whole company
Fernanda Oi treats making a research team visible as part of the work, not an afterthought.
Read more
They present the team formally, show the organisation who is on it and name the people involved rather than claiming the achievement alone.
They name João Maia as the person who created the team, credit the marketing team for the announcement format and thank individual team members by name.
They also describe a long-standing difficulty accepting that career wins are the product of their own effort, a tension alongside their habit of deliberately sharing credit.
They wrote and assembled a company-wide announcement email for the UX research team from a car under time pressure.
The marketing team produced the HTML, and they only later noticed that the email carried their own digital signature.
The moment landed at a monthly company meeting, where their name, their colleagues' names and the heading "Equipe de UX Research" appeared together.
They describe an embarrassed grin of pride.
A later post states their current position, UX Research Manager at Venturus, thanks João Maia and Antonio Salgado for staying with them, and names the team: Maria Luiza Gregório Dias Nery, Vinicius T. and Patricia Queiroz.
The two posts trace a shift from doing research to managing people who do it: one introduces the team, while the other offers a year-end statement about the manager role and the people it now leads.
What is discovery, in Fernanda Oi’s own words, and where does research fit?
Read more
Start here to see how they frame the problem a product team is trying to avoid.
Fernanda Oi announced an article co-written with Antonio Salgado about Discovery, one of the services in Venturus’s Design portfolio.
The post states its purpose plainly: Discovery aims to make a product or service more likely to hit the mark by finding problems before development rather than after.
They say the work applies UX Research and UX Design techniques, not as a fixed recipe but with each engagement customised to match the challenge to a solution quickly.
The post points readers to the article and invites them to make contact.
For brands planning a campaign
For a company planning a discovery phase, this names a service with stated goals — assertiveness, custom scoping and speed — that can go into a brief or be compared with other firms’ offerings. For a consultancy selling similar work, it shows how a research-led discovery offer is positioned for buyers: not as a fixed package, but as work scoped to the challenge.
Which tool should a research team reach for when it needs depth on a deadline?
Read more
This post answers with an actual test, the prompt used and the criteria that separated the two tools.
Fernanda Oi ran the same prompt through OpenAI’s Deep Research and Gemini’s Deep Research, asking for behavioural data on a defined user group — single people with children — drawing on scientific articles and practitioner sources in Portuguese and English from 2020 to 2025, and correlating findings across neurobiology, behaviour, psychology and cognition.
OpenAI’s results fell short: sources outside the requested period, including one from 1987; reliance on sites that cited research rather than the research itself; and a shallow final text that drew no connections between the data.
A colleague suggested trying Gemini, which produced a structured, editable research plan from the same prompt, with closer adherence to the brief, relevant sources, deeper analysis, academic sources and a list of sources considered but not used.
They extended the test with three separate reports on coparenting from different angles — parents, children and external influences such as family and support networks — to see whether the tool would repeat itself or generate genuinely different material.
The reports were consistent without duplicating content and offered complementary insights.
A final cross-report analysis surfaced patterns and outliers and referred back to the earlier reports without collapsing into a summary.
The post concludes that ChatGPT is not automatically the best or only option, that Gemini still stalls at times and produces over-long text, and that research efficiency means going deeper without losing time or quality.
For brands planning a campaign
A vendor of AI research tooling can respond directly to the weaknesses identified here — date drift, secondary citations and no synthesis across sources — while the post offers a citable, prompt-level comparison for content about research tooling. A research team evaluating tools gets a reusable test design: the same prompt, explicit date and language bounds, then a follow-up to check whether a second and third report add anything new.
What does it take to stand up a research function inside a company?
Read more
This first-hand account covers building a team and presenting it to an organisation, including the work beyond research methods.
Fernanda Oi describes presenting a UX Research team to a company and writing and assembling the announcement email under time pressure, including work done from a car, while the marketing team produced the HTML.
They had not realised that the email, written on behalf of the team they were coordinating, carried their digital signature and would reach the whole company.
The moment landed at a monthly company meeting, when their name, their colleagues’ names and the heading ‘Equipe de UX Research’ appeared together.
They name João Maia as the person who created the team, credit the marketing team, thank team members Vinicius T., Juliana Mariano and Maria Luiza Gregório Dias Nery, and thank Venturus for the opportunity.
They also note a long-running difficulty in accepting career wins as the product of their own effort.
For brands planning a campaign
Useful for employer-brand, internal-communications or talent-attraction briefs where the story is a function being built, not just an individual’s expertise.
The post also shows how Fernanda Oi shares credit: they name colleagues and the person who created the team rather than claiming the work alone.
That matters when planning a collaborative or co-branded engagement.
Where is Fernanda Oi in their career, and who do they work with?
Read more
This short post establishes their current title, employer and team.
In a reflective year-end post, Fernanda Oi says they are now UX Research Manager at Venturus, after a year they describe as asking for challenges and then finding plenty of them.
They thank João Maia and Antonio Salgado for supporting them and name their team: Maria Luiza Gregório Dias Nery, Vinicius T. and Patricia Queiroz.
The post marks a transition from doing research to managing people who do it.
For brands planning a campaign
Confirms seniority and managerial scope for briefs that need a manager to sign off on or resource a project, and indicates that an engagement would likely run through a team at Venturus rather than one individual. The post also names people who appear alongside Fernanda Oi elsewhere, offering context when assessing who owns which part of a programme.
Testing AI deep-research tools against real UX questions
Fernanda Oi ran the same behavioral-research prompt through OpenAI's Deep Research and Gemini's Deep Research and compared the outputs.
Read more
OpenAI returned sources outside the requested 2020–2025 window, including a 1987 article, relied on secondary sites that only cited studies, and produced a shallow synthesis that did not correlate findings across neurobiology, behavior, psychology, and cognition.
Gemini returned an editable research plan, stayed closer to the prompt, used academic sources, and listed additional relevant sources it had not used.
Oi then stress-tested Gemini with three linked reports on coparenting and asked for a cross-report analysis of patterns and outliers; the result offered a new reading rather than a summary.
The stated downsides: the tool sometimes freezes, and the texts run long.
The same prompt was run in two tools, followed by repeat report requests on one theme and a separate synthesis pass to check for duplicated content and complementary findings.
For brands planning a campaign
For a B2B SaaS brief on product discovery, this shows a hands-on approach to selecting and pressure-testing AI research tools before using them in a study, with a documented account of where they fail.
Fernanda Oi co-authored an article with Antonio Salgado presenting Discovery as one of the services in the Venturus design portfolio.
Read more
The stated aim is to improve a product or service's chances of success and avoid problems after development. The article presents customized UX research and design—not a standard playbook—as a way to match a specific challenge to a solution quickly.
Customized UX research and design techniques for each engagement, rather than a repeatable template.
For brands planning a campaign
A discovery-phase brief can define which user questions to answer before development starts and why standard templates were set aside for the challenge.
Coordinating a UX research team and raising its profile
Fernanda Oi describes introducing a UX research team to the company by email and seeing the team and its members named at a company-wide monthly meeting under the title "Equipe de UX Research."
Read more
In a later post, they identify as UX Research Manager at Venturus and credit João Maia and Antonio Salgado as ongoing supporters, along with team members Maria Luiza Gregório Dias Nery, Vinicius T., and Patricia Queiroz. The posts show both the coordination work and the deliberate effort to introduce the team to the wider organization.
Internal communication and recognition: formally introducing the team and naming its members at a company-wide meeting.
For brands planning a campaign
For a brand that needs discovery research run through an internal team, this shows experience recruiting, coordinating, and raising the profile of a research function—not just running studies alone.
The posts identify Venturus as Fernanda Oi's employer and mention two AI research tools tested in their work, but do not disclose whether any sponsorship, affiliate, client or investment relationship exists or provide post dates.
Read more
Ask for disclosure and post dates.
Venturus
Employment
Fernanda Oi says they are UX Research Manager at Venturus and write about coordinating the UX Research team, including introducing it to the wider company.
Read more
They also promote a Discovery article about a service in Venturus's Design portfolio, co-authored with Antonio Salgado.
At a colleague's suggestion, a post compares Google's Gemini Deep Research with OpenAI's on the same prompt.
Read more
It credits Gemini with an editable research plan, sources within the stated date range, academic references and three complementary reports, which they then asked it to cross-analyse.
Fernanda Oi publishes mostly in Portuguese on LinkedIn.
Read more
Start with the post comparing OpenAI's Deep Research and Gemini's Deep Research on a real behavioural-research question, then read the announcement of a co-authored Discovery article with Antonio Salgado and the posts about building and presenting a UX research team at Venturus. The recurring questions are which AI research tools earn their place, how discovery is scoped to a specific product challenge and how a research function becomes visible inside a company.
Publishes reflective LinkedIn posts that trace progress over time, name the people involved and connect individual work to a team or portfolio outcome.
Read more
Cynthia Schatzmann focuses on freight logistics, broker relationships, customer success and driver retention; Fernanda Oi focuses on UX research, discovery and AI research tooling.
Writes practically about how AI and data projects move inside companies: the quality of inputs, tool adoption and explaining change to colleagues.
Read more
Both use short LinkedIn posts to test an idea against a business setting.
Rafael Soares focuses on data governance, CIO communication and enterprise AI adoption; Fernanda Oi focuses on UX research tool selection, product discovery and research team visibility.
Shares structured B2B operating advice for managers, breaking a process into before-and-after responsibilities and asking readers to apply it to their own team.
Read more
Each post makes a clear argument and invites readers to comment.
Diego Lorenzo Carvalho writes about sales management, Sales Ops, CRM pipeline reviews and prospecting as founder of SalesX; Fernanda Oi writes about UX research, discovery and AI research tools.
Fernanda Oi publishes on LinkedIn to a small professional following: 2,161 followers as of 9 October 2026.
Read more
For a B2B SaaS buyer, that is a targeted professional audience rather than a broad-reach placement.
A product-discovery campaign aimed at people who commission or run research may fit the topics covered—UX research, discovery work, team coordination, and an AI research-tool comparison—but the four self-published posts do not show who reads them or provide impressions, engagement rates, or reader job roles, seniority, and employers.
Ask Oi for recent post-level results—impressions, reactions, comments, and reposts—on the UX research and discovery posts closest to the campaign, plus any audience survey or platform data on reader roles and seniority.
Do not use follower count alone to set expected campaign reach.
Fernanda Oi's four LinkedIn posts cover UX research practice, product discovery services, team leadership, and AI-assisted research methods.
Read more
They do not describe sponsored formats, rates, deliverables, or a commercial inquiry route.
The commercial context they do establish is service-based: discovery work delivered through Venturus's design portfolio, customised to the client's product challenge rather than offered as a packaged media placement.
Buyers should confirm directly with them or Venturus whether a partnership is available and what terms apply.
Strong fit for a B2B SaaS company looking to improve product discovery.
Read more
Fernanda Oi co-authored a Venturus article that presents Discovery as a design service: using custom UX research and UX design, rather than a fixed recipe, to make a product or service more likely to succeed before development.
That aligns with campaigns focused on improving product discovery.
The posts also show a practical research method under deadline: define a user group, set source-date and language bounds, compare tools, then check whether additional reports add findings or repeat existing ones.
As UX Research Manager at Venturus, they have coordinated a research team, making this a fit for someone who can both run and resource discovery work.
Reach and commercial format remain unverified: the LinkedIn account had 2,161 followers as of 9 October 2026, with no impressions, engagement, job-role, seniority or employer data.
Ask directly about paid terms and available formats; the posts do not describe sponsorship options or an inquiry route.
Plan this as an expert collaboration or service engagement around discovery, not a broad-reach sponsored placement.
When to choose a specialist
Poor fit for a broad consumer awareness campaign that needs large reach and measured audience delivery, such as a mass-market DTC beverage launch targeting general consumers.
Read more
The Portuguese-language professional LinkedIn account has 2,161 followers, with no audience demographics or engagement data.
The content covers UX research, discovery services, research team management and AI research tools—not consumer lifestyle, retail, entertainment or mass-market topics.
The posts provide no sponsorship offer, media kit, rate card or inquiry link, so buyers cannot plan a packaged paid placement from the available information.
For a mass-reach campaign, ask directly about audience delivery and paid formats; neither is established by the posts.
A worked campaign example
Good fit for a B2B SaaS company seeking a research plan to improve product discovery and looking to draw on Fernanda Oi's UX research and discovery expertise.
Campaign angle: a Discovery-before-build teardown.
They take a real or anonymised product concept from the company and break down the discovery plan vendor and client should agree on before development: which user questions must be answered, which method is customised to the challenge rather than taken from a template, and how the team will know the answer is strong enough to proceed.
Use the AI deep-research test as a working segment.
Show the exact prompt used in the post: a defined user group; scientific and practitioner sources in Portuguese and English from 2020 to 2025; and correlation across behaviour, psychology and cognition.
Then show the checks that separate a usable output from a shallow one: sources outside the date window, secondary sites that only cite studies, duplicated findings across reports, and whether a cross-report pass surfaces patterns and outliers.
The deliverable could be a co-authored article, a recorded working session or a short resource for product and research leads, with Venturus credited as the service context.
Ask Fernanda Oi or Venturus whether the engagement runs through Venturus Discovery, what the team can deliver and what commercial terms apply; the posts do not specify these.
Pick Rafael Soares when the discovery question turns out to be a data-foundation question.
Read more
His posts argue that AI and analytics work succeeds or fails on the quality and organisation of data — information spread across folders, servers, separate clouds, print, email and social channels — and that many companies lack the team or time to fix that before buying tools.
If a B2B SaaS company cannot trust its own data enough to research or automate against it, that readiness conversation is closer to the real blocker than a research sprint.
He also writes about the internal side of the same rollout: his shared CIO article describes explaining, educating and training employees who fear being replaced, which matters when a data or AI capability has to be adopted across teams.
The tradeoff is that they do not document their research or discovery methods.
Neither post shows how they frame user questions, test concepts before build or run studies, so a campaign aimed at improving product discovery would be buying adjacent data work and still need research expertise elsewhere.
Both posts read as service outreach rather than close readings of client work, leaving little public method to evaluate.
Pick Diego Lorenzo Carvalho when the decision shifts from product discovery to commercial execution — specifically, how a manager finds the cause when a sales team misses quota.
Read more
His posts make the same argument Fernanda Oi makes for research, applied to pipeline: the person closest to the problem should arrive with their own analysis and a named root cause before anyone proposes a fix.
His longer post maps a change in the sales manager's role, from owning the number and checking deals one by one toward developing specific skills and conversion metrics with each seller, and having sellers bring their own read of the funnel numbers into pipeline reviews.
That is a usable angle for a vendor selling sales-enablement or revenue-operations software to the same B2B SaaS buyer.
The tradeoff is subject: this is sales management, not product or user research.
Neither post covers discovery, user testing or design services, so this material does not answer the question of how to improve product discovery.
He identifies himself as founder of SalesX and writes to sell that service, making the posts commercial outreach.
Pick Cynthia Schatzmann when the brief changes to growth and retention in a relationship-driven market with a bilingual US audience.
Read more
One post describes rebuilding a broker portfolio through daily, direct communication, using dispatcher experience to plan where drivers should be positioned for the best loads, and adding two partnerships on her own initiative — reporting roughly 241% growth in monthly load volume and 158% in gross freight over three months.
A second post covers six months at a logistics company, entering an unfamiliar field and a second language, supporting 11 drivers, and reaching $200,000 in delivered loads, published in Portuguese and English.
For a campaign that needs first-hand stories about customer success, retention and selling into US freight — rather than product discovery — she has the more directly relevant material.
The tradeoff is topical distance: none of the posts concerns research, product discovery or software design, so Fernanda Oi would not serve a campaign aimed at improving product discovery.
Their material consists of self-reported milestone and percentage claims without underlying figures, and their strongest content is personal results narrative rather than a method a reader can apply.
Ask for audience size, sponsorship format and an inquiry route.
Four public LinkedIn posts by Fernanda Oi are in Portuguese, and no publication dates appear in them.
Oi's LinkedIn profile showed 2,161 followers as of 9 October 2026.
The page includes overview, ideas, relationships, discovery, and campaign-fit sections.
The Discovery article announced in one post, along with any books, shows, or earlier work, is not detailed here; identity remains an open gap, and complete name-search coverage is not claimed.