Investor data room · Lyra AI, S.L. · Barcelona

Every hire, explained.

A software company for AI hiring. Our software helps companies find the right people, and tells them why. We sell it through forward deployed engineers (FDEs): our own engineers, who join a customer's team for a while and build AI tools there.

€6Mpre-money valuation
€500kpre-seed round
1,200+women in our community, 50+ countries
€0 → €10kMRR in three months, from services

Inside this room

Eighteen chapters. Start with the people, then the business and the research, then the numbers. Every figure in the model is yours to change.

    Forbes 30 Under 30 España 2026United Nations ITU AI Skills CoalitionCruz Roja EspañolaBarcelona Supercomputing Center AI FactoryParloa, design partnerConforamaAI Learning Lab with Anthropic and Make Forbes 30 Under 30 España 2026United Nations ITU AI Skills CoalitionCruz Roja EspañolaBarcelona Supercomputing Center AI FactoryParloa, design partnerConforamaAI Learning Lab with Anthropic and Make

    01 Founder story

    We are the workers this product is for.

    Stephany Oliveros, founder and CEO of Lyra AI

    Founder & CEO · Product and research

    Stephany Oliveros

    Founder, AI researcher and product lead with 7+ years of experience. University lecturer in AI at EU Business School. Forbes 30 Under 30 España 2026, recognised for helping women find jobs with Lyra AI and for her work in education. Founding contributor to the UN ITU AI Skills Coalition, one of Spain's 100 women referents in tech, speaker at MWC and 4YFN. Co-founded the women in AI community that feeds Lyra AI and leads our partnerships with the Red Cross and the Barcelona Supercomputing Center.

    Kristina Talova, co-founder and COO of Lyra AI

    Co-founder & COO · Operations

    Kristina (Kris) Talova

    Founder. Project manager and educator with 10+ years leading complex programmes. Expert in linguistics applied to education and learning design, studying educational frameworks for AI. Designed and ran the educational programmes that reached 1,200+ women globally, and builds the operations behind our cohorts, content, placements and partnerships.

    A note from Stephany and Kris

    We are two migrants who built our careers in tech in a second language, in a country that was not ours. We know what it is to be filtered out before a person has read a word.

    A CV that does not fit the template. A career break that reads as a gap. A background that a keyword filter does not recognise. When AI arrived in hiring, we watched it learn exactly those shortcuts and apply them at scale, with a confidence score and no explanation.

    So before we wrote a line of product, we built the people side. Two years teaching women AI: 1,200+ members in 50+ countries, 500+ women trained in AI literacy in the last ten months, roughly 40,000 people reached through talks. Companies started asking us for the people we trained.

    Lyra AI is what we built when the answer had to be more than an introduction. It is personal for us. It is also the right business: when anyone can generate a CV and copy an algorithm, the scarce things are verified people and trust. We have spent two years earning both.

    Stephany Oliveros & Kris Talova

    Readiness

    Proof before promises.
    Ready before day one.

    30U30Forbes España 2026, for helping women find jobs and for our work in education
    €0 → €10kMRR in three months, ~40% margin
    UNFounding contributor, ITU AI Skills Coalition
    1,200+women in the community, 50+ countries
    Red CrossSigned collaboration agreement with Cruz Roja Española
    BSCIn residence at the Barcelona Supercomputing Center AI Factory
    500+women trained in AI literacy in ten months
    <7 daysto place our first two beta hires, both women
    3companies founded between the two of us
    Anthropicand Make sponsor our AI Learning Lab: 20 seats, 100 on the waitlist

    02 Company story

    A community first. Then a company.

    We started with people. The software is what those people, and their future employers, asked for.

    1. 2024

      The community

      Stephany co-founds a women in AI community and runs the first AI literacy cohorts. It grows by word of mouth across 50+ countries.

    2. 2025

      The education engine

      Kris designs the programmes that make the teaching repeatable. Talks and workshops reach roughly 40,000 people. Companies start asking who they can hire.

    3. Apr 2026

      Lyra AI, S.L. is incorporated

      Barcelona, CIF B27660489, Via Laietana 26. The thesis: verified people plus explainable matching.

    4. May 2026

      Institutions say yes

      Residence at the Barcelona Supercomputing Center AI Factory. Collaboration agreement with Cruz Roja Española. University partners follow.

    5. Summer 2026

      Services pay the bills

      Forward deployed AI engineering and placements take us from €0 to €10k MRR in three months at about 40% margin. Conforama trusts us with AI implementation.

    6. 2026

      Forbes 30 Under 30

      Stephany is named to Forbes 30 Under 30 España, recognised for helping women find jobs with Lyra AI and for her work in education.

    7. Oct 2026

      Parloa and Vega

      Parloa, one of Europe's fastest-growing AI companies, joins as a paying design partner. Our beta places its first two hires, both women, in under 7 days. The pre-seed opens.

    03 The problem

    If you don't play the game, you're not in.

    I have always hated psychometric games and ATS (Applicant tracking system) filters. They do not find talent. They find people who are good at passing filters.

    The best hire I ever made was a waitress. I hired her for a customer success role. She had not been to university, had no LinkedIn, had never worked in an office and could not make a PowerPoint. She had everything that is hard to teach. She became head of customer success. No ATS would ever have shown her to me.

    Kris and I are the same story. We had the least linear careers possible and none of the usual opportunities, so we kept proving ourselves. I grew up with nothing, in the middle of the mountains, and ended up building an international organisation that works with the United Nations and changed education for thousands of women around the world.

    We are building what would have found us: a way for companies to find exceptional people that the current funnel throws away, with a reason for every decision.

    Stephany Oliveros, founder and CEO

    4%of applicants who applied to ten roles were rejected by all ten, more than chance predicts

    The evidence

    Algorithmic monoculture is real.

    Stanford researchers studied about 4 million applications screened by one vendor's algorithm, pymetrics, between 2018 and 2022. When many employers rely on the same few screening vendors, the same people get rejected everywhere: 4% of applicants who applied to ten positions were recommended for rejection by all of them. They also found adverse impact on Asian and Black applicants under US employment standards.

    Europe has decided this is not acceptable. Under the EU AI Act, hiring systems are high-risk and must be transparent, auditable and overseen by people. Most of today's tools were not built for that. We are setting the new standard instead of retrofitting to it.

    Bommasani et al., "Algorithmic Monocultures in Hiring", Stanford Digital Economy Lab, 2026
    1. 01

      The ATS is useless now

      Everyone uses AI to tailor their CV to the keywords. Filters read the same polished document a thousand times and learn nothing about capability.

    2. 02

      Games are not talent

      Psychometric games reward people who have learned to play them. Non-linear careers, career breaks and unusual backgrounds are filtered out quietly, at volume.

    3. 03

      Everyone gets the same algorithm

      When thousands of employers buy the same few screening tools, one bad score follows a person everywhere. That is monoculture.

    4. 04

      Nobody can defend the shortlist

      Scores are fast and unexplained. In Europe that is now a legal problem as well as a quality problem.

    Who it is for, and what they get

    04 The product

    Vega: AI agents to hire with confidence.

    Give Vega a role. Get a shortlist of up to ten people with reasons and evidence, outreach sent and interviews booked. Vega recommends; a person decides every step. From 2027 it runs entirely from Slack.

    app.justlyra.com / roles / ai-engineer

    Pipeline overview

    How we get the data

    The Life CV: a profile you earn, not one you write.

    A CV says what someone claims. The Life CV is built from how they actually answer, calibrated before their first applications and refined with every role.

    1. 01

      Sign up

      A short questionnaire of job preferences, and a CV upload that we parse into skills, roles and tools.

    2. 02

      The AI interviewer

      Before the first few applications, the candidate answers short, spontaneous questions by voice memo. Unrehearsed answers are hard to game.

    3. 03

      Calibration

      The first answers calibrate the profile; each later answer refines it. A few questions at the start, then less and less.

    4. 04

      Themes, not transcripts

      Transcripts go through thematic analysis. We keep the themes with tone-of-language tags and delete the transcript. No audio is stored and no biometric voice analysis is done.

    5. 05

      The company side

      Companies go through a similar assessment of how their teams really work, so both sides are described in the same language.

    6. 06

      Affinity constellations

      The engine maps the two sets of themes against each other. Every link is a reason a recruiter can read.

    Try it

    Affinity with this company·

    Illustrative data. Themes light up as evidence builds; links appear only when both sides support them.

    Try it: four scores, each explained

    Every score is 0 to 100 with a minimum bar, a plain reason and at least one evidence pointer. A reason with no evidence is rejected by our validator and shown as "not enough evidence". Move the weights the way a recruiter would.

    Candidate 4F2A Anonymous until mutual interest

    Forward Deployed Engineer · Barcelona · hybrid

    0%match

      AI recommendation. A person makes the decision. Protected traits are never inputs to scoring.

      05 Go to market

      Distribution, guaranteed before launch.

      Two motions. Forward deployed engineers (FDEs) get us inside companies, where we ship and see the hiring need first. Strategic partnerships bring talent, credibility and institutions. The talent side is the easiest part for us: we have grown it for two years.

      What is an FDE?

      FDE means forward deployed engineer. It is one of our own engineers.

      The FDE joins a customer's team for a few months. They build an AI tool the customer needs, like a support agent or a hiring workflow.

      The customer pays us every month for this, from €4,500 to €6,500. One FDE works with two customers at a time.

      While they work, the FDE sees who the customer needs to hire. That is how the customer starts using our software, Vega.

      1,200+women already in our community, across 50+ countries, before launch. Our first supply channel; the marketplace itself is open to everyone.
      500+trained in AI literacy in ten months: the first verified talent cohort
      30,000our talent pool target, converting the community and partner universities into talent users

      Motion 1

      Forward deployed engineers

      1. Land. A blocked engineering leader with an AI role open for months. We ship one working system in 21 days.
      2. Expand. Retainer, then the next workflow. We become the team that knows their stack.
      3. Convert. When they hire, they hire through Vega and our marketplace. Every engagement adds verified evidence to the supply.

      Today: Conforama trusts us with AI consulting and implementation. Parloa, building its own FDE organisation in Europe, is our paying design partner. With Deel, clients can hire the people we verify anywhere, remotely and compliantly.

      Motion 2

      Strategic partnerships

      06 The company

      A software company. We sell it with engineers.

      Lyra AI makes software that helps companies hire the right people. That is the business. To win customers, we first send our own engineers to work inside them. The customer pays us for that work. They learn to trust us. Then they buy the software.

      1. 1

        We send one of our engineers into a company. We call them an FDE.

      2. 2

        The FDE builds something useful with AI. The company pays us every month.

      3. 3

        The company sees our work up close. Now they trust us.

      4. 4

        When they need to hire, they use our software, Vega.

      5. 5

        Every hire teaches Vega something. It gets better for everyone.

      Vega · AI Engineer

      Candidate 4F2A86%
      Candidate 91BC78%
      Candidate 2D0771%

      Vega: our software

      This is our main product, and what the money is for. A company tells Vega who it needs. Vega finds the right people and explains why. The company pays every month.

      • Our main product
      • From €400 a month
      • Up to €4,000 to €10,000 a month for big users

      21-day sprint

      Day 14 of 21

      Support agent live in staging

      FDEs: how we sell

      An FDE is one of our engineers, working inside a customer. It gets us in the door. It also pays from day one. One engineer works with two companies.

      • How we win customers
      • €4,500 to €6,500 a month
      • Two companies per engineer
      Agency desk · our supply
      4F912DA7

      Agencies: our software, their desk

      Recruitment agencies find people for companies. Many use old tools. We set up our software for them. They pay once to start, then every month to use it. Even our competitors can buy from us.

      • About €30,000 to set up
      • Then a monthly fee for use
      • Our talent, inside their tools

      Life CV

      AI literacy 88

      Gap: production LLM evals · 2 lessons

      Job seekers: free to join

      People looking for work join for free. They see what they are good at, and what to learn next. Later, they can pay for courses. We do not count that money in our plan.

      • Always free to join
      • Free learning links today
      • Paid courses later

      How we find and win customers is in Distribution.

      Who we verify

      Tech talent first. Then every department.

      The edge is the same in every role: they understand AI, and we know it from our own assessments, not from what a CV or a self-rating claims.

      Now
      • AI engineers
      • ML engineers
      • Data scientists
      • Forward deployed engineers
      • QA
      • Full-stack engineers
      From Q2 2027
      • Strategy
      • Sales
      • Marketing
      • Operations
      • HR
      • Finance
      • Legal
      Evidence
      • AI interviewer calibration
      • Role-specific AI literacy tests
      • Shipped work
      • Upskilling completed
      How the four parts reinforce each other The community feeds FDE engagements, which produce cash and verified evidence that powers Vega, whose customers and agencies hire from the marketplace, which grows the community. Community1,200+ women, 50+ countries FDE engagementscash and verified evidence Vegacompanies and agencies Marketplaceverified, explained supply One company,one thesis.

      We do not read CVs. We watch people ship. Today we place those people inside companies. Next year we open that proof to everyone who hires.

      07 Investors

      The round is moving. Top-tier investors are in the room.

      We are in active conversations with top-tier VCs and angels in the United States and the United Kingdom. A €500k round has limited room. If you want a place in it, talk to us early.

      United States

      Top-tier VCs and angels

      In active conversation about the pre-seed.

      United Kingdom

      Top-tier VCs and angels

      In active conversation about the pre-seed.

      08 Market opportunity

      The largest staffing market in the world is learning to hire AI.

      We start where the pain is sharpest and the price is highest: senior AI, ML and forward deployed roles in Spain and the EU, then the UK and North America.

      AI is changing how every company hires. The volume of AI roles is rising faster than the supply of people who can prove they do the work.

      The line between "recruiting" and "engineering" is dissolving: the buyer who needs an AI engineer wants the system shipped, not a stack of CVs.

      The numbers tell their own story:

      €221BEuropean staffing market in 2025, the world's largest region: about 39% of a €566B global market. SIA
      +1,165%forward deployed engineer job postings, January to October 2025 against the same months of 2024. Bloomberry
      24.8%annual growth of AI in HR, from $3.25B in 2023 to $15.24B by 2030. Grand View Research

      Bottom-up, stream by stream

      Our own assumptions, anchored to public counts. Change them, then compare with what each scenario in the financial model needs.

      Market signals

      Why now

      Three shifts meet in the same 18 months.

      1. AI talent is the bottleneck. Forward deployed engineer postings grew more than twelvefold in 2025. Companies cannot ship AI because they cannot find the people.
      2. Agents are finally good enough to know both sides deeply: the candidate's whole story and the company's real working style.
      3. Regulation forces explainability. EU AI Act transparency duties apply since August 2026; high-risk obligations for hiring systems land on 2 December 2027. Black-box screeners must rebuild.

      Why later is too late

      Matching software is becoming a commodity: a USD 2.5bn HR suite just bought an AI matching engine to bundle as an add-on. What cannot be bought is a trusted, verified supply of people and the distribution to reach them. That takes years and it compounds. Whoever owns it when the deadline hits sets the standard for Europe.

      09 Why choose us

      They have solved speed. None has solved trust.

      We skip black-box scores and unverified databases to build hiring that a manager, a candidate and a regulator can all read.

      Everyone else

      Automate the funnel as it is: score AI-written CVs faster, pull from databases nobody verified, and leave the hiring manager with a number they cannot explain.

      Lyra AI

      No ATS needed. Scores come from complete, verified data, not AI CVs. Talent we trained and assessed. A matching engine that reduces bias by design and is auditable from day one.

      Competitive positioningHorizontal axis: owns a verified talent supply. Vertical axis: explainable and compliant by design. Lyra AI sits alone in the top right.

      Honest assessment

      Where they win, what worries us, and what we do about it.

      From our own competitor research. The category is funded and crowded, and it will get more so. The model assumes that.

      "You can reverse engineer any algorithm. You cannot reverse engineer a marketplace, a distribution channel built over two years, or trust."

      10 Architecture

      Technology that explains the score.

      Not a ChatGPT wrapper. Our matching engine uses graph theory to connect each person's universe to the data map of a company, and computes every match as a sum of named, weighted contributions. Language models read and write. They never decide.

      Graph theory, in one picture

      A candidate is a set of nodes: skills, shipped work, preferences, learning. A company is another: must-haves, team, working style, goals. Matching finds and weights the edges between them. Each edge is a sentence a recruiter can read and a regulator can audit. Our logo is that graph.

      
      
      
      
      
    
      
    
      Why this technology, and not an LLM
      
    
      
      
      
        
      LLM wrapperLyra AI engine
      
        
          
      Personal dataRaw CV text sent to a third-party APIStructured profile and embeddings stay in our EU database
      
          
      ExplainabilityA number with a plausible paragraphPer-dimension contributions, each linked to evidence
      
          
      AuditabilitySame input, different answerDeterministic and replayable; model, weights and version logged
      
          
      Bias controlHidden criteria learned from the internetProtected traits stripped before scoring, enforced by tests
      
          
      Cost at scaleAn API call per candidate per roleVector and graph search, near-zero marginal cost
      
          
      EU AI Act, Annex IIIRebuild requiredAudit trail, human oversight and documentation by design
      
              
      
      
    
      
    
      • Next.js and TypeScript
      • Supabase Postgres, EU region
      • pgvector
      • Edge functions
      • Claude, for extraction and explanations only
      • Mistral Voxtral speech to text (EU)
      • Slack app
      • WhatsApp, email, Cal.com
      • Row-level security on every table
      
  
      

      



      
  
      
    
      
      
      11 Research
      Agents that learn from each other. The research behind the product.
      
      
      Lyra AI is a research company as much as a product company. We work with the Barcelona Supercomputing Center AI Factory, led by a computational neuroscientist.
      
    
      

    
      
      
      Why we own the algorithm
      Independent of big tech, by design.
      
      
        
        
      • We own the data. The matching runs on our engine and our EU database. Candidate data is never training data for someone else's model.
        • 
        
      • Privacy by construction. Themes instead of transcripts, no audio kept, protected traits never scored.
        • 
        
      • Prices that hold. Language models only read and write explanations, and can be swapped for open models. If OpenAI, Anthropic or Google raise prices, change terms or close access, our margins and our service stay.
        • 
        
      • Built for a downturn. If the AI bubble deflates or another financial crisis hits tech, a deterministic engine on our own infrastructure keeps running at near-zero marginal cost.
        • 
      
      
    
      

    
      
      
      
        
      The hypothesis
      
        
      Humans are social creatures. We learned not only through evolution but through each other: trade, exchange, culture. Social learning.
      
        
      Today's AI agents work from one view of the world: their system prompt, their own feedback loops, and data someone chose for them. DeepMind built open-ended virtual worlds where agents receive no instructions and new, general behaviour emerges from play.
      
        
      An intelligent, self-sufficient creature raised alone in the wild becomes Tarzan. Let it learn socially and it becomes something more evolved than the original.
      
        
      So what kind of agent emerges if a company's agent, with its corporate world view, and a talent's agent, with an individual's, share one virtual world and trade behavioural cues?
      
        DeepMind, "Generally capable agents emerge from open-ended play", 2021
      
      
      
      
        
      
          
      
            
            
          
      
          
        
      
        
        
      
          
      Behaviours found0
      
          
      Hybrid agents0
      
          
      Running0s
      
        
      
        
      Company agents Talent agents. An illustration of the question, not research results.
      
      
      
    
      

    
        
      
      1. 2027
        Sandbox
        Company and talent agents built from consented, anonymised Life CV and company themes, in isolated and shared worlds at the BSC AI Factory.
        2. 
      
      2. 2027 to 2028
        Measure
        Do agents that trade cues find more strategies, and better matches, than isolated ones? Preprint, then peer review.
        3. 
      
      3. 2028 onwards
        Apply
        What holds up feeds assessment design and matching. What does not, we publish anyway.
        4. 
    
      
  
      

      



      
  
      
    
      
      
      12 Financial projections
      Run the numbers yourself.
      
      
      A quarterly operating model from Q1 2027 to Q4 2029. Opening cash is the round only: the product is 100% funded by it, and what services earned before 2027 stays with the service side. Pick a scenario, then change any assumption. Hiring choices in the Team chapter feed the same model, and the roadmap reads its numbers quarter by quarter.
      
    
      
    
      
    
      
    
      
      
      
      
        
      Revenue by line, per quarter
      
        
      Cash and EBITDA
      
        
      Show the full quarterly table
      
        
      Inputs are management assumptions, not forecasts of fact. ARR counts recurring revenue only (FDE retainers, Vega, agency usage credits, B2C); software ARR excludes the FDE retainers. Implementation fees and commissions are one-off. Vega and agencies are tracked by signing cohort: new customers start at the entry price and expand with usage. New software customers are capped by sales capacity. FDEs are payroll from the day they join, bill from the month after, and work at the stated utilisation. Customers pay after the stated number of days (Spanish law caps B2B terms at 60). Salaries and FDE prices are indexed +3% each January. Costs include overhead per person, recruiting per hire, a fixed platform cost, and corporate tax at 15% for the first two profitable years, then 25%, after loss carry-forward, paid the following July. Not modelled: VAT timing, the two extra Spanish salary payments in June and December, depreciation, grants and debt. The Deel partnership is not in the numbers.
      
      
      
    
      
  
      

      



      
  
      
    
      
      
      13 Interactive roadmap
      A platform, a research asset, then a global pipeline.
      
      
      Click through the next twelve quarters. Milestones are our plan; the numbers come live from the scenario you chose in the model.
      
    
      
    
      
    
      
    
      
    
      
  
      

      



      
  
      
    
      14 Team and hiring plan
      A small team that ships, with advisors who have done it.
      
    
      
    
      
      
      Hiring plan
      Switch a hire on or off to see its effect on burn, runway and breakeven. Everything recalculates in the financial model.
      
      
        
    
        
  
        

        



        
  
        
    
        
      
        15 Chat with us
        Ask us anything. Pick who answers.
        
      
        We both know the whole company, but we each go deeper on different things. Steph designs the product and thinks about where it goes. Kris runs the day to day and knows every number. Choose one, ask the questions every investor asks, or your own.
        
    
        
    
        
      
        
        
        
        
        
          About youTell us your fund and thesis, and we answer for your portfolio
          
        
            
            
            
            
        
            
            
        Kept only in this browser and used to tailor the chat's answers.
        
          
        
        
        
        
        An AI version of each of us, answering only from this data room and the live financial model. For anything binding, or anything it cannot answer, write to hello@justlyra.com.
        
      
        
      
        
        
        StephProduct, tech, research and the roadmap
        
        
        
        
        
        
        
          
          
          
        
        
      
        
    
        
  
        

        



        
  
        
    
        
      
      
        
        
        16 The round
        
        
        €500k pre-seed at a €6M pre-money valuation.
        
        
        The valuation does not come from software alone. It comes from the team, the research, the proprietary technology, what we have already done, the market and a model you can stress yourself.
        
      
        
    
        
    
        
      
        01
        The team
        Three companies founded between the two founders, a community built from zero to 1,200+ in 50+ countries, a CPO who was an engineering manager at Factorial, a lead engineer, a computational neuroscience research lead, and advisors from Amadeus, a Microsoft acquisition and legal.
        
      
        02
        The research
        A research programme with the Barcelona Supercomputing Center on agents that learn socially. A potential standard for how hiring AI is built, not a feature.
        
      
        03
        Proprietary technology
        Our own matching engine and the Life CV data behind it. Independent of any model vendor, auditable for the EU AI Act, near-zero marginal cost.
        
      
        04
        Experience
        Revenue from services from month one, our first two beta hires placed in under 7 days, Forbes 30 Under 30, the UN, the Red Cross, Conforama and Parloa. We have sold, delivered and been paid.
        
      
        05
        Market opportunity
        The world's largest staffing market, a twelvefold jump in FDE job postings, and a regulation that forces every hiring tool in Europe to be rebuilt by December 2027.
        
      
        06
        Projections
        A bottom-up model that assumes competition grows every year. Even the conservative case reaches about €1.7M ARR by the end of 2029; the model is open for you to stress.
        
    
        
    
        
      
        
        
        How €6M compares
        
        
        Not fitted to a spreadsheet. Below the market median.
        
        
        We did not pick projections to justify the price. These are public benchmarks, and the live multiple from whichever scenario you choose in the model.
        
      
        
      
        
        
        Our round: €6M pre-money, €6.5M post-money
        €6.5M post-money, roughly $7.5M
        
        
        Median post-money SAFE cap, pre-seed raises of $250k to $500k
        $10M Carta, Q2 2025
        
        
        Median seed pre-money, AI applications
        $20M Carta, Oct 2025
        
        
        Post-money ÷ ARR end 2027, this scenario
        ·
        
      
        
      
        Carta data covers mostly US companies and skews to AI; European rounds price lower, so treat these as a ceiling, not a target. Converted at about 1.15 USD per EUR.
        
    
        
    
        
      
        
      
        
      
        What the first 18 months cost, and who pays
          
    
          
  
          

          






          
  
          
    
          
      
          18 FAQs
          Short questions. Short answers.
          
      
          The questions investors ask us most. Numbers come live from the financial model, so they change with the scenario you choose.
          
    
          
    
          
  
          

          



          
  
          
    
    
          
      
      
          Get to know us. Then let's build it.
          
      
          hello@justlyra.com · Via Laietana 26, Barcelona
          
      
          Book a call with the foundersjustlyra.com
          
    
          
  
          
  
          Confidential. Shared for the sole purpose of evaluating an investment in Lyra AI, S.L. Forward-looking figures are estimates. Not an offer of securities.