Free
€0
Post a role and see your first match, at no cost.
- Post jobs for free
- Scan 50 people for free
- Get your first match
- 15% success fee, only after a hire stays three months
In development
Vega: every shortlist, explained.
Vega is Lyra’s hiring agent, named after the brightest star in Lyra. Today, our team supports sourcing and screening, with a recruiter reviewing every match. We are building the remaining stages around one shared brief and explainable recommendations.
Every score comes with its reasoning.
Each candidate is assessed on four axes. Every score has a plain-language explanation, and the candidate sees exactly the same panel as the hiring manager.
Candidate A
Senior platform engineer, remote EU
89
Average of the four
-
Experience 87/100
Shipped a payments SDK and two internal design systems. Meets all eleven required criteria; Kubernetes, a preferred skill, is missing.
-
Soft skills 81/100
Led a team of five for three years. Gave structured answers in the screen and explains trade-offs clearly.
-
Cultural fit 95/100
Prefers async work and written decisions, as your team does. To discuss: wants to meet in person every quarter.
-
AI literacy 92/100
Trained in AI with Lyra. Uses agents daily and knows where a human should stay in the loop.
What works today, and what is next.
One brief carries through every stage. Each stage shows where it stands.
-
Live now
Sourcing
Candidates come first from Lyra’s network of women trained in AI, then from verified external sources. Most are mid-senior, with a smaller pool of juniors from reputable universities. Every name comes with a one-line reason.
-
Live now
Screening
A shortlist of up to ten ranked names, scored on the four axes, each with its explanation and a link to the evidence. A recruiter reviews every match.
-
In development
Intake
The hiring manager describes the role in Slack or the web dashboard, by voice note or text. Vega asks what a good recruiter would (team size, must-haves versus nice-to-haves, salary band, working style) and writes the brief.
-
In development
Interview prep and scoring
Questions and a rubric aimed at each candidate’s gaps. After the interview, a recommendation with strengths, risks and how the view changed since screening: a draft for the panel, not a verdict.
-
Planned
In-interview notes and onboarding
Notes taken against the rubric during the call, then a 90-day plan and a team-context brief for the new joiner and their manager.
How the matching engine evolves.
Built in the open, one phase at a time. A recruiter verifies every match today; each phase adds signal and automation without giving up the plain-language reasoning.
-
Live now
MVP: manual concierge
A recruiter verifies every match by hand.
-
In development
V1: vector embeddings
Cosine similarity with hybrid metadata filters.
-
Planned
V1.5: career graphs
Graph theory across skills, roles and trajectories.
-
Research stage
V2: agentic matching
Self-improving, multi-signal and Slack-native.
Human review and explainable recommendations.
Vega is built so a person can read and defend every decision: each score is explained, candidates can see and correct what is said about them, and a recruiter reviews every match today. One rubric is applied the same way to every candidate, so the bar does not move from one review to the next.
Skills are evidenced, not claimed: portfolios, situational judgment tests and courses on the platform. Matching on evidence rather than keywords widens the pipeline to people screening filters usually hide, particularly women returning to work, career switchers and self-taught engineers.
Under the EU AI Act, AI systems used for recruitment are classified as high-risk; the rules for these systems apply from 2 December 2027. Our approach draws on research in fairness in ranking, explainable AI for high-stakes decisions, bias auditing in hiring (including the EU FINDHR work) and graph representation learning for careers. We aim to take this work to FAccT, RecSys and the EU AI Act technical standards work.
Pricing.
The success fee is charged only after a hire has stayed three months, so the placement guarantee is proven before you pay. Stages still in development open to each plan as they ship.
Growth
€350 a month
More reach per role, no success fees. For small companies and recruiters.
- Unlimited job postings
- 100 scans per role
- 10 matches per job posting
- No success fees
Vega credits
€200 a month, metered
Run Vega, our AI recruiting agent, priced per credit: from 200 to more than 30,000 credits.
- Everything in Growth
- Interview prep, generated per role
- Interview booking, handled end to end
- Onboarding analysis after every hire
- Dedicated support and onboarding
Straight answers.
Is Lyra a recruitment agency?
No. Vega is a hiring product in development. Today, our team runs sourcing and screening with a recruiter reviewing every match; that human review is also how the matching engine learns.
Can I use Vega today?
Sourcing and screening are available as a supported service. Tell us about the role you need to fill and we will explain what is possible today.
Who is on the candidate side?
Lyra’s network of women in tech, trained in AI with us, plus external candidates. Candidates see the same scores you do.
What does getting started look like?
A 30-minute scoping call, then a calibration session that maps your roles, success criteria and current high performers to the four scores. Your first private shortlist follows within 10 working days, and nothing is charged before it lands.
Do I have to use Slack?
No. Slack is the fastest way in, but a web dashboard sits underneath and can drive the same pipeline.
Have a role to fill?
Join the waitlist, or tell us about the role and we will explain how the current service works.
Join the Vega waitlist