Most recruitment processes are not designed. They are inherited. A job description gets copy-pasted from two years ago, a five-step interview flow survives because nobody had time to question it, and a hiring team quietly doubles in size while the process behind it stays exactly the same. Inês Frade, co-founder of AryaHub, sees this pattern constantly across the tech companies she works with. In a recent conversation, Inês walked through the methodology AryaHub uses to help companies scale recruitment without chaos: why AI is not the fix people hope it is, and what actually breaks once a company crosses the 20-30 person mark. What follows is a case study of that methodology, built entirely from her own words.
TL;DR
- Recruitment processes that work fine at 10-30 people tend to break once more stakeholders and decision-makers get involved.
- AryaHub diagnoses before it implements: mapping stakeholders, decision-makers, and RACI ownership before touching a single tool.
- Change is rolled out gradually and co-designed with the teams who will use it, to raise adoption rather than force it.
- AryaHub’s embedded recruiter model (RPO) places consultants inside client teams, where they run daily operations and flag structural fixes as they go.
- AI is treated as a tool for consistency and speed, not a fix for a broken process. Automation is reserved for repetitive, low-value tasks.
- Unmapped skills inside a growing company are described as “a pile of gold” that most leadership teams cannot actually see or use.
- AryaHub’s own next challenge: applying its own methodology internally as it grows past 20-30 people itself.
Why Recruitment Breaks at 20-30 People
AryaHub was founded two years ago by Inês Frade and Rui Miranda, both coming out of consultancy and startup backgrounds in tech HR. Inês previously helped establish Parsly’s hub in Portugal; Miranda was leaving Teamlyzer. The idea behind AryaHub was to stop treating recruitment as an isolated function and instead cover the full HR life cycle for tech companies.
Inês has a clear read on when things start to crack. Below roughly 20 to 30 people, informal communication carries a company through most process gaps. Past that point, more people need to be involved in decisions, and if authority is not clearly assigned, everything slows down or stalls.
“Usually until 20-30 people, everything works because you can speak with proximity, but then after that, decisions need to be made, and if you’re stuck within one decision maker, things start to break. When more stakeholders come into the place, usually it starts to break as well.”
The specific symptom Inês sees most often in recruitment is a job profile that never got updated. A description written when the company was ten people and hiring a “senior Java developer” keeps getting reused long after the actual need has shifted to something like a senior DevOps engineer. Nobody rewrote it, so nobody notices the mismatch until interviews start going nowhere.
The Diagnosis-First Methodology
AryaHub’s answer to this is a rule the company holds to deliberately: understand before you implement. Inês describes clients arriving with a process that quietly stopped matching how the business actually operates, and the first move is never a new tool or a new workflow.
AryaHub maps out stakeholders, decision-makers, and the company’s RACI structure (who requests a task, who decides, who runs day-to-day execution) before proposing any change. Many companies, according to Inês, do not have this mapped out at all and cannot say clearly who owns what.
Only once that diagnosis is complete does AryaHub build a roadmap and move into implementation. The reasoning is practical: a tool bolted onto a process that was never re-examined tends to just make the existing errors happen faster.
Why Gradual Change Management Wins
The second discipline in AryaHub’s approach is pace. Inês’ view is that implementation only holds if it is introduced gradually, because organizations tend to reject too much change injected at once. The roadmap gets agreed with leadership, then shared with the teams who will actually use the new process, with those teams involved in redesigning it.
“You mentioned a pretty good point, which is change management. We always do a roadmap that is of course aligned with leaders, but then is shared with the teams, and we usually bring the people to help us redesign the processes and accommodate new changes, in order to increase the acceptance rate.”
This matters because software adoption failures are rarely about the software. A tool gets rejected when it does not fit a process the team never agreed to change in the first place. Bringing the team into the redesign is what gets the acceptance rate up before a single license is purchased.
Embedded Recruiters: The RPO Model in Practice
A large part of AryaHub’s work is what Inês calls the “embedded recruiter” model, more formally recruitment process outsourcing (RPO). A recruiter, junior to mid-senior depending on client need, joins the client’s team directly and runs daily recruitment operations alongside them.
Unlike a standalone freelance recruiter, an AryaHub embedded recruiter follows the company’s Kaizen-based continuous improvement methodology: identifying friction in the day-to-day process and proposing fixes as they surface, rather than waiting for a scheduled review.
Inês gave a concrete example from a client where AryaHub now has four embedded recruiters. Two weeks into the engagement, one recruiter produced a report showing that hiring in Porto, a small and specific talent market, would require a higher budget than planned after contacting 400 candidates and analyzing the response rate. The client used that data to shift the role from on-site to fully remote and shortened its interview process from five phases to three.
Companies exploring the CV management challenges consulting and staffing firms face at scale will recognize the pattern: the fix rarely starts with a tool. It starts with someone close enough to the process to see where it is losing time or candidates.
AI Is Not a Magic Pill
Inês is direct about where AI fits and where it does not. Dropped onto a broken process, AI does not fix anything, it scales the existing errors faster.
“AI is not going to replace HR people, it’s only going to replace the people that don’t use it well and that don’t know how to implement it. If you throw it on top of a broken process, it will just scale the errors that the process already has and will not bring clarity.”
AryaHub’s sequence is to map the client’s process first, then decide, task by task, whether AI or automation is the right fit, since Inês points out the two get conflated constantly despite doing different jobs.
AryaHub uses AI internally for interview note-taking and CV formatting. Automation gets applied to low-effort, repetitive tasks such as scheduling interviews or sending candidate invitations, freeing recruiters to focus on strategy and candidate relationships.
The same principle shows up in how Marco Pincho described Sprint CV’s own approach during the conversation: converting interview notes directly into a formatted CV, translating between languages without altering content, and letting recruiters add only what changed rather than retyping a candidate’s history from scratch. It is workflow automation built around a specific, already-understood process, not AI applied as a blanket fix. Companies weighing whether this kind of automation is worth building can start with the AI CV Parser or the broader case for a CV management system.
The Hidden Cost of Unmapped Skills
One of the sharper points in the conversation concerns skills visibility. In a ten-person company, everyone knows what everyone else is good at. Past a hundred or two hundred people, that knowledge disappears unless it is deliberately captured.
“If you don’t have a process or a framework that allows you to access those skills and competences, you’ll be losing, because you don’t know what your people are good at.”
AryaHub introduces frameworks that help managers assess team skills and help individuals understand their own career direction, aiming for a match between what the company needs and what motivates the person. Inês’ framing is that without this, a company is sitting on what she and Marco both called “a pile of gold” it cannot use, whether that gold is unassessed internal talent or, as Marco pointed out, a database of past candidates nobody revisits.
Common signs a company has lost visibility into its own recruitment process
What This Means for Consulting and Staffing Firms
The conversation also touched on a scenario specific to consulting, staffing, and firms bidding on public tenders. When a large account needs to respond to an RFP or procurement process asking for, say, fifty Java developers with specific utilities experience, recruitment teams frequently stop recruiting altogether just to locate and format the right CVs.
“I can tell you by knowledge that a large company’s recruitment team stopped recruiting and stopped doing things because they were five days allocated preparing CVs.”
This is where Inês’ skills-mapping point and Marco’s CV automation point meet directly. A company that has already mapped its people’s skills, and already has a system for generating compliant, formatted CVs on demand, avoids losing a five-day chunk of its recruitment capacity every time a tender lands. Firms evaluating this gap can look at how other enterprise teams have approached it or work through the Enterprise API for connecting CV data directly into existing systems.
- Do you know, without checking, how many people in your company have 5+ years of a specific in-demand skill?
- If a tender landed tomorrow asking for fifty formatted CVs in a specific template, how many days would that take?
- Can a hiring manager tell you who owns the final decision on a new hire, without asking around first?
- Has your core job description template changed in the last twelve months?
- Do your recruiters know which repetitive tasks are eating the most time in their week?
Frequently Asked Questions
What is an embedded recruiter?
An embedded recruiter, also called RPO (recruitment process outsourcing), is a recruitment specialist who joins a client’s team directly to run day-to-day hiring operations, rather than working as an external freelancer. AryaHub’s embedded recruiters also bring continuous improvement practices from other clients into the engagement.
How do I know if my recruitment process needs to be redesigned?
Common signs include job descriptions that have not changed in over a year, unclear ownership of hiring decisions, recruiters spending hours manually formatting CVs, and a candidate database that never gets revisited for new roles.
Is AI replacing recruiters?
According to Inês Frade, AI is not replacing HR and recruitment professionals. It is replacing the ones who do not learn to use it well. Applied to a broken process, AI amplifies existing errors rather than fixing them.
What is a RACI map in recruitment?
RACI stands for Responsible, Accountable, Consulted, Informed. It is a framework for identifying who requests a task, who makes the final decision, and who runs the daily operation. AryaHub uses it during the diagnosis phase before recommending any process change.
Why do recruitment teams stop hiring during tender season?
Large staffing and consulting firms responding to RFPs or public procurement often need dozens of formatted, compliant CVs on short notice. Without a system to generate them quickly, recruiters manually reformat CVs for days, halting active hiring in the process.
What is the difference between AI and automation in HR processes?
AI supports judgment-based tasks such as writing, note-taking, or maintaining a consistent tone across communications. Automation removes repetitive manual steps, such as interview scheduling or candidate invitations, entirely from a person’s workload.
Final Thought
Nothing in AryaHub’s approach depends on a specific tool. It depends on sequence: diagnose before implementing, involve the team before changing their process, and only then decide where AI or automation actually earns its place. That sequence is also why the conversation lands as relevant well beyond recruitment specifically, since consulting, staffing, and tender-driven businesses run into the same gap between what a company knows about its people and what it can actually act on. AryaHub’s own next test, as Inês put it, is applying that same discipline to itself as it grows past the very headcount threshold where she says things start to break.
About the Author
Three years of turning Sprint CV’s workflows into tutorials, videos, and guides people actually finish taught António Teixeira what separates a feature nobody adopts from one that sticks. He is part of the team that writes the guides, walkthroughs, video tutorials, and blog content that help recruitment and consulting teams get more out of the platform, translating every product update into content clients can act on immediately. Connect with António on LinkedIn
