Why We Built Hireo: A Behind-the-Scenes Look at BetterQA's Recruitment Platform

Every software product has an origin story. Hireo's begins with frustration.

BetterQA is a software quality company with 50+ engineers spread across multiple time zones. We test software for clients in fintech, healthcare, and defense—industries where quality isn't optional. To do that work, we need to hire exceptional people. And hiring exceptional people with existing tools was painful.

The Problem We Faced

In 2023, our recruitment process looked like an efficiency nightmare spread across four distinct stages, each one more time-consuming than it needed to be.

The first stage was simple enough: receiving CVs. Applications arrived via email, job boards, and LinkedIn in different formats and structures—some in PDF, some in DOCX, and occasionally a JPEG of a handwritten CV (yes, really). Getting applications wasn't the problem.

The second stage was where pain began: manual data entry, consuming 15-20 minutes per candidate. Someone had to open each CV, extract contact information, work history, skills, and education, then type it all into our ATS. Formatting inconsistencies meant constant copy-paste-fix cycles that drained productivity.

Skills assessment came next, adding another 5-10 minutes per candidate. Reading through each CV to tag relevant skills raised constant questions. Does "Java" mean they actually know Java or did they just list it? Does "5 years experience" in their summary match their actual work history? Manual verification was required every single time.

Finally, job matching took 30+ minutes per role. When a new project came in requiring specific skills, someone had to manually search our candidate database, filter by keywords, and review each potential match individually.

The math was brutal. With 100 candidates per month at 25 minutes of processing each, we were spending 42 hours monthly just on data entry. Add 10 new roles per month at 30 minutes of matching each, and that's another 5 hours. Factor in context switching, interruptions, and human error, and we were burning nearly 50 hours monthly on tasks that felt like they should be automated.

Why Existing Tools Failed Us

We tried the alternatives. Enterprise ATS platforms gave us powerful features we didn't need and weak features where we needed strength. They had great compliance workflows but terrible CV parsing. Startup ATS tools had beautiful interfaces and shallow functionality—their AI features were marketing copy, not reality.

AI-powered screening tools could parse CVs but couldn't integrate cleanly with our workflow. Standalone tools meant double data entry. Manual processes with spreadsheets worked initially but collapsed as we scaled past 20 candidates in the pipeline.

None of them solved the core problem: turning unstructured CVs into structured, searchable, actionable candidate profiles—quickly and accurately.

What We Built

Hireo started as an internal tool. The requirements were simple and driven by our daily frustrations.

We needed to parse CVs in seconds rather than minutes, using AI that actually understands document structure. We needed to extract skills intelligently, recognizing that "React.js" and "ReactJS" and "React" are the same thing. We needed to calculate experience accurately by deriving years of experience from work history rather than relying on self-reported summaries. We needed to match candidates to jobs semantically, understanding requirements rather than just keywords. And we needed to generate branded CVs on demand—when presenting candidates to clients, we required professional, consistent formatting.

We spent six months building and refining. Our QA engineers—the same people who test mission-critical software for banks and defense contractors—tested Hireo relentlessly.

The Engineering Decisions

Several key architectural choices made Hireo different from alternatives.

AI-Native, Not AI-Augmented

Most ATS platforms bolt AI onto existing workflows. "Now with AI!" stickers on features that were designed before AI was practical. We built Hireo assuming AI would handle the heavy lifting from day one. The data model, the workflows, the entire architecture assumes intelligent automation. When you upload a CV, AI isn't an optional add-on—it's the foundation.

Parse First, Ask Questions Later

Many systems require you to fill in forms, then optionally attach a CV. We inverted this approach. Upload a CV, Hireo extracts everything, and humans review and correct only what needs correction. This reduced data entry time from 15 minutes to 30 seconds per candidate.

Skill Taxonomy, Not Keyword Soup

We built a comprehensive skill taxonomy that maps variations to canonical entries. "JavaScript," "JS," "ECMAScript," and "ES6" all map to JavaScript. "Machine Learning," "ML," "Deep Learning," and "Neural Networks" are mapped and related appropriately. "5 yrs," "5 years," "five years," and "2019-2024" all normalize to numeric values. Now you can search for "Python" and find candidates who wrote "python," "Python3," or listed it implicitly in project descriptions.

Branded CV Generation

As a services company, we present candidates to clients. Every agency recruiter knows the pain of reformatting CVs into consistent templates. We built CV generation into Hireo's core. Click a button, select a brand template, get a professional PDF. There's even an option to anonymize for blind screening. CV presentation dropped from 20+ minutes to 30 seconds.

Results After One Year

Internal metrics after running Hireo for our own hiring told a compelling story. CV processing time dropped from 15-20 minutes to just 30 seconds—a 98% reduction. Skills tagging went from manual to automatic, eliminating that task entirely. Candidate matching went from 30+ minutes to instant. CV presentation dropped from 20 minutes to 30 seconds—another 98% reduction. Data accuracy improved from around 80% to approximately 95%.

Our recruiting team's capacity effectively tripled without adding headcount.

Why We Made It Public

Internal tools often stay internal. Hireo didn't, for two reasons.

First, other companies have the same problem. Every conversation with clients revealed identical frustrations: manual data entry, keyword matching that misses good candidates, time wasted on formatting CVs. The problem is universal, and we'd already solved it.

Second, usage improves the product. More users means more edge cases, more CV formats, more feedback. Our AI improves with volume. Making Hireo public accelerates that improvement for everyone, including us.

What We Learned

Building Hireo taught us things we now apply across all BetterQA projects.

AI needs human guardrails. Hireo's AI is impressive, but we designed it to surface uncertainty. Low-confidence parses get flagged for human review. The goal isn't replacing humans—it's amplifying them.

Quality compounds. The same rigor we bring to testing client software went into testing Hireo. Edge cases, error handling, performance under load. Quality isn't a feature; it's a practice.

Eating your own cooking works. We use Hireo daily. Bugs get found immediately. Friction gets felt personally. Product decisions come from real experience, not abstract user research.

Where We're Going

Hireo is production-ready and being used by teams beyond BetterQA. We're focused on expanding AI capabilities with better parsing for non-English CVs and more sophisticated matching, building an integration ecosystem connecting to job boards, HRIS systems, and communication tools, adding multi-language support with full localization for European markets, and developing enterprise features including advanced permissions, audit logging, and compliance workflows.

But the core mission stays the same: make recruitment intelligent, not tedious.

Try It Yourself

Hireo started as a solution to our own problems. If you're facing similar challenges—manual CV processing, keyword matching that misses good candidates, time wasted on formatting—try Hireo free for 14 days.

We built it because we needed it. We're sharing it because you might need it too.

Michael Torres is Product Lead at Hireo and Engineering Manager at BetterQA. He's hired 50+ engineers using the screening methods built into Hireo.

About BetterQA: BetterQA is a NATO-certified software quality company serving clients in fintech, healthcare, and defense. Our 50+ engineers test mission-critical software where failure isn't an option. Hireo is one of several internal tools we've built and made available to others—because good tools shouldn't stay locked inside one company.