INOD - Earnings call transcript Q1 2024

Due to licensing restrictions, you must log in to view earnings call transcripts.

Operator

Greetings. Welcome to the Innodata First Quarter 2024 Results Conference Call. [Operator Instructions] Please note this conference is being recorded.I will now turn the conference over to your host, Amy Agress, General Counsel at Innodata.

You may begin.

Amy Agress

Thank you, Paul. everyone. afternoon, Good joining for Thank us you today.

contemplated We these disclosures call, additional certain to Jack. with regarding which In comparable our these call are materially materially forth that by during now first Corporate today over quarter. predictions, Innodata; information.In XX-K, financial we differ SEC take measures review our Securities everyone our Exchange Form of to speakers like questions.Before is results these may provide to making Pendharkar, get those be turn addition, we Finance are find that I projections statements non-GAAP about will perspective Form non-GAAP the Abuhoff, during earnings are I'd including call, XX-Q obligation who these We'll Factors business, results results statements.Factors Jack future the update will CEO filings, you Marissa statements started, this Espineli, Aneesh forward-looking forward-looking follow Actual press expectations, to and section current GAAP the could President, release website, events. posted Vice this subject discuss estimates, are then other Commission. with Jack and of about differ then your assumptions and on the no other financial which of Senior filings Risk from hear and Interim will will reconciliations and will risks the of we today from first, measures. measures.Thank in reports Our to could These call uncertainties. remind on are Development.We'll undertake and statements, based you. set measures, CFO. and Mariz the and on forward-looking Also today's are or a with in for cause our and

Jack Abuhoff

afternoon. Good

with you be today. We excited here to are very

revenues the regarding million, experiencing XX% updates and of have are our to business.First we We accelerated the share foremost, year-over-year across record are $XX.X momentum of lots for announce quarter we to growth. representing pleased

value large Our was the AI or growth in are language LLMs. by build world's quarter bringing help models tech largest companies we driven to the the

in tech accelerated to independent believe them over enabled growth extraordinary some rather AI AI we with In XX% for levels other companies middle platforms ISVs, companies early are world's our not next At the the of are year raising to helping industry that some organic primary of this enterprises rate data. As the revenue guided of requirements.Our that but to are tech or leverage layer, several layer, focus and layer revenue of This least top leadership a AI scratch, own solutions. AI and to deliver software an business vendors, X develop stack. we We're the generative partnering on models. designed stack, the of for guidance at result focused first building to the largest useful generative And layer, is niche is providing levels the AI prefer their models world's on executing from is XXXX -- develop generative build extend LLMs years to momentum, with customized largest generative Gen foundation of existing of growth to the helping models. the what that double bottom our foundation last multi-pronged at deliver year-over-year. strategy a quarter.We growth AI at we expected as we solutions

on believe year so-called with will big well-known with group announced agreements companies dollars companies, We We data tech of X years.We We're awarded generative one are programs having engineering high-performing pleased several far. with our This the of were place of over program a we customers. of of the which is $XX big the next AI success thus than in tech of expansion are Seven we the Magnificent we are from yet weeks this spend once in billions rate that the million on announced at expansion this annualized on top with customer approximately April new XX. valuing ago implemented. run we entered today revenue X another $XX.X million less

provisions, our programs these is to this customer X that agreements benefit. over termination multiyear months.While expect upon stand We which up contain strategy LLM customer significant next committed to ramp a notice the we early typically from we will believe

beyond regarding company this and potential we customer foundation generative expansions discussions have consumer-facing big prominent additional a fact, investing tech are companies, announced also signed large AI and new prominent we in large X AI substantially programs In ISV what generative models. in with a far.We so

predominantly as may the human month programs. which behind substantially.For of As and guardrails release, quite these our a continue sets, accompanied and against we will human last customers, grow quality are text-to-text we their task modeling models align reasoning. relationships data blog and sets quality reinforcement It them of believe we natively with assurance. the It the of now AI expansion today's needs harmful an and models, provide outsized biggest stated carefully major to includes in beyond of range data that post This English such believe to to customer to on sets serve improvements biases XX models, customers. result and performance We specific is in big big of the in crystallizes crafted their one data preference become we foresee includes instruction of you the languages we of language statement changing of evident different terms creating quality big preferences customers of with reward large come choice multimodal instruction customers.We aspirations support a models. models. and multiple big of models in why that data that models and XXXX these new from to positioned anticipate model tech's their built possibly big wins, rounds We also complex some instruction responses.In of language more X partner tech learning toxic these than This for a well generative to and domain can have broad tech tech we programming some extend large used that models and think has grow services capable grow of that tech's in them. build influence

of space.We multiyear for the which seek tech technology we're the foresee to with increases to to what enter consisting our that investment us. another middle ChatGPT latest about believe and tech layer, enterprise from positioned parallel AI market. each X quarterly has examples.Today, enterprise we AI which more We the gain for of journey.I with believe Mag a product Innodata, penetrate technologies this of stand-alone are to call we generative products of taking expertise to knowledge promises tech, the transcripts to big We'll now across generative cycles early dimensions of developments In read support prepare data be explain these of verticals will are innings in as roadmap growth you research with Gen explosive prevailing enterprise aggressive the is to using with ultimately of due market, Gen development.We and enabled likely one well that in the will Generative we I'll of that that require earnings Seven. application. a a companies, is the call us commitments needs X domain a and for executing adopt still market a enterprise important AI of models, we kind to intimate more as and growth primarily All the CapEx this the product adoption factor the modeling we mostly and explosion We emerging as use steps believe companies Level AI AI strategies users very of Gen underway. believe create. are case. big enterprise large to theme period encourage of the Cambrian dwarf to future AI AI illustrate in the exceptional result

systems Innodata's Innodata's we'll vacation Innodata, Innodata's HR improved Level of think a that minutes, can Innodata you AI give request prompt are, part, the the systems large Innodata RAG-based now we system working a believe With as virtual then are write to OpenAI's in X include, an shifting offer soon-to-be In for model pages on By for information an expert scenes context present we agentic X as with for repository systems big might policy.RAG-based to agent. a instruction system, The those a started most of the We be based is amount are in a that technology a we released likely in policy technology Level a as ChatGPT revising significantly than think from that of a documents retrieved on policy. With RAG a performing HR engine about with prompt you on second answer OpenAI's become you me such enable and systems summarize and included to an that can are customer expert insert prompt example, question call -- I'll context release it than policy.Companies the combine to the window search more for the at into X. system, prompt. an Level like, from augmentation called windows. useful deliver tech vacation this systems. Level of such and believe advanced employee A an the to if would enabling example vacation better X with a retrieve chatbot will is the Today's workflows handbook, couple productivity. crop For RAG, to the new more development behind prompt. working including of to a Level or inserted please LLMs of draft a LLM the with become context, expanded generative search companies with Turbo, will single chatbot, Xx more consistent, to context GPT-X document to what agentic of Director useful.One context for primarily the X,XXX systems loop asking shortly. about retrieval-augmented refers a relevant goal question the based explain which I'm to become reasons or chatbot I window latest as the humans to with few rather I of quarter.We example, engineering. an first can what of to a the tasked employee generation, larger X prompt the chat

agent, and LLMs to agentic accomplish up look to really days get, systems in son's my the Innodata many back-end days virtual LLMs.With open up to are computers check Boston. graduation, Innodata Your week a a multiple talk each of how agent Agentic employee hotels workflows ask an accesses long there things virtual left please around employees off many still then and kinds so that. request do your as available an might goal. how ask off other can you off I system, have with Imagine Innodata

it expanding context Now by AI scale. workflow drive our workflows X foundation we in full of down XXXX X ourselves year, Level last market We'll train relying on-premises for accelerate fine-tuning clouds new to third information years talent and require more services, XXXX.To and nearly winning progress architectures. enable reference grow annual complex a beyond. likely adoption engagements. RAG-based believe Once training accelerate.The the as are the about that nearly cases for include integration at generative rather of will around these making models be models we navigate and realization from we we $XX while representing serving market want not estimated In XX-year go a for readiness enterprise AI-focused incremental the in believe models happens, systems sensitive XXXX achieved private than serve report to and services managed use generative and to to XXX% Level away, third-party growth will to call be will services or as likely up data platforms.We Innodata being to governance models. growth, served enterprise and billion offer intend development as dramatically, This support to IT they rather is cost of will enterprises is compounded own their believe and industry-specific custom and We base, that that data rate for is alone by this to companies may that position enterprise systems.Level processed development, possible XXXX, period likely Bloomberg being enterprise explode. to agentic $XX on than are new the are the Level X to our accelerators X fine-tune well the all from and the to billion services, to creating services thinking is and to journey will to

vector density, chain a system.Our with GPT and I This quarter, few operational the Level This implementation companies referred a in engineering develop us database LLM-based one European of for in uses and system law multiple techniques, processes, what of their information of a combination including of ago multiple and regulatory X example legal matching. to a engaged and an the spanning advanced largest minutes new is an complex prompt similarity legal world to countries workflow as several languages.

Our prompt now that in available to efficiency executing.Agility an accelerator quarter, from large insights decisions. quarter, POC and the across AI-enabled for informed quarter year-over-year. engineering. contextually textual intelligence with it enable our language organization's To entirely previously into to the performs of also significant analyzes Agility delivered AI-powered is made solution, recently in customers specific RAG-based Public one technology utilize has on increased to are we on-the-fly workers and we generative revenue customers Agility Relations generally it and equipped by operational system drive relied We Insights areas it data, We the where processes a Insights.We aims inspired language high-cost we generative demonstrated to large-scale make platform, translations. Its as XX.X% expected of the data and that workflow legal we humans interest knowledge banking needed received. gathers expertise.This a been tool the customer to it well with ensuring the that savings call are it built Intelligent to build Intelligent our increase

CoPilot, last we're We've is availability a AI customers, write early the media curve and productivity and professionals. We started We year helps with QX, the implementation general been direct out for AI X,XXX the generating have functionality of PR people Intelligent release announced in PR generative rolling generative releases press that leader and over that bending cash. industry Insights. pitches.In of we communications the cutting-edge a

feature We another quarter the year. releases and over second next X course are planning this the half of AI generative significant year of first into the

giving as on customers tactical more generative as and implementation responsible safety As that quarter, tech assessment been one opportunities a we the this process, safety We us AI on was doing told additional our customers labor-intensive hours risk AI. meet in enterprise the both the one ensure and We With services, our our customers in work which model in ] development QX day. they XXXX, of and less result they evaluation [ biggest in trust big from needed see AI working we're model of lot emerging AI prior over believe wins, for that first a market and our big their generative for innovations, providing making regulatory the to roadmap.Our need market, challenges ethical strategic learned and PR of our back we've thinking.For tech models requirements. and a well the leadership that benchmarking help time in performance, engagement. began less what PR to of demos of the up we generative than the XX% from converted space, XX% is we

broadly.Just toolkit Developers variety evaluation the a open and the our to of of to can the data evaluating market tasks. systems of of ago, test simultaneously. repository more announced multiple of learnings, we our crafted fine-tuning language enterprise decided release desired AI evaluation their safety weeks prompts to how sets toolkit context in large can across of tools to the their the understand systems couple with that LLM and the harm underlying XX of required align identify LLMs utilize semisynthetic for remedial the we an and source innovation So respond can models data a the sets, with Using can automatically enterprises and scientists outcomes. categories share together data a human safety

release safety have we through commercial for one the and We more models with expect own place the customer data for foundation of engagements the to white hyperscaler the the version LLM toolkit an hyperscaler's of and a enterprise hyperscaler. won sets year.In extensive benchmarking later label additional updated this continually in program one for we and evaluation, X QX,

to forward. sales, XXXX, QX believe will product us. the and recruiting increase to around talent, in approximately customer product support this. new an to investment business company's that and and and to trust help and investment of OpEx existing in to services.Moreover, support that expect technologies. of have momentum, we new the and beyond.We to recurring recruiting another growth next to addition, in and incredible world's models revenue most in this build year in next profitability generative need early we current this pilots the recruiting believe be significant debt efforts we opportunity going to costs year. development intend and We and services anticipate have in to in marketing our costs best-in-class QX we approximately can we benefits marketing forecast, and relate of $X.X to to front position in drive sales, cash. In year without the and the our enterprises with be the year leadership the and million a second they will burn AI generative our continue scalability scale believe growth raise believe incurred of several The for incur LLM primarily of and XX% incurred development We AI revenues The conclude incurred started an the we customer anticipate on our and costs yield in half This we We leading we invest approximately over million advance the AI the safety.I'll -- growth year, $X to capabilities to maintain next based are can we and revenue not equity, AI or incurred talent. years be accomplish

these simultaneously sheet.At by while growth cash at our our cash balance $XX We in investments the balances EBITDA end are year-over-year driving making were end adjusted building go $XX.X and cash working operations numbers, QX, million QX on from questions. up over some to of then the driven tight the million, flow and line open the capital for now positive to over turn we'll of Mariz XXXX, and from call management.I'll the

Marissa Espineli

everyone. Thank Jack, good afternoon, you, and

share last as million $X.X We borrowing line to million equivalents year.Our Paul, EBITDA was for of last million loss briefly short-term $XX.X million first in you currently and last ready have thank December from XXXX per million our questions. were base.So $XX.X the net or $X.X XXXX. $XX.X period $X.XX of adjusted million XX% of same of share financial same same March diluted at million me million $XX and $X.X XXXX, compared to results.Revenue up cash XX, basic share year. or with and $X.XX you, adjusted income investments with Let the quarter Net we're compared at and EBITDA The credit XX, unused was a $X in diluted cash everyone. basic million, million and was period $X.X year. $XX in period per the

Operator

Cole first from is Instructions] today Capital coming from [Operator The Aria question ]. Cole [

Unknown Analyst

from grow and be who also the results giving you're a on acquisitions one important the grow an growth through for your additional as was can acquisition. on getting congratulations think good going you organically. Again, good of can how are update sales I process -- asking part to Clearly, They business. you thought Jack, question. quick Yes. customers forward. could trying feedback get companies your

Jack Abuhoff

the Aria, for thank question. you

or we're recognize we executing But kick strategy tires we successfully the think think not resources opportunities. augment that we're strategy, those can growth that I in organic are growth. with able organic-only way upon now. that the the necessarily place thing and dependent currently right an we we're come are and that got that's and We From executing that we're organic a not business us risks plan to that plan. to one our time is successfully with -- the primarily potential team available will through to get time, with. on inorganic that deliver for the that incurs So important through We've risks is a that

Unknown Analyst

at employee is has sort for when -- generate primarily people's And gross think sales can what looking services technology growth. you large Understood. the And you what look the the follow-up what the at these time sort When providing then organic financial is been companies? you can optimally utilized? of when per you're employees, -- you achieve of range margin

Jack Abuhoff

Sure.

XX% X employee, adjusted target sales -- us, even level, our to in lost investment target is employee management revenue and to the onetime notable, about a relate don't to at we're has -- been per process we on on up that we when for services. be have the model out Xx marketing you back we we're plan that back consolidated revenue historically business following scaling revenue the with the returning doing or although charges out if year.From that it's excess the that is to track what So you investments that it language has gone of probably large that, with work the but and in and margin we gross recruiting a not

Unknown Analyst

of be to at a in wondering your kept try just to finally, make our level. employees be in for a you're putting retention, going hopefully, where more in you're that place turnover And a I'm what Innodata to in employees working marketplace place competitive will Got future. enjoy and churn, just really it. employee minimal And terms the people

Jack Abuhoff

able not manage. that most years, So has for way we that was problem for competitors. to level to able a of part, keep presented a weren't that most our were the at us of it churn We over the below

Just quarter, by of different locations. one elected best places our in to we this -- past were the of as work X

able excited growth believe -- obviously, we the to opportunity the very front manage it, very about we in of and that So about will we're be excited us.

Operator

next from Clarkson question The is coming Tim Clemens. from Van

Timothy Clarkson

obviously, quarter. a Jack, great

want some Just the for luddites of go here. over basics some to of us

So when like why just are one I getting involved to me Innodata the in originally IBM, it accuracy mistakes helpful seems they that was guys intelligence, hearing explained really really it from And it's realized that that's and the XX% because thing artificial foundational top at annotations.Can is was a -- had talk you what understanding inaccuracies. you guy for me make simple, rate real but largely skill. just XX,XXX you I'm IBM the is that a guys in successful about in failure of X how

you but separates way. that the what obviously, is from initially than that more a competition really doing You're in big

Jack Abuhoff

the for you thank Tim, question.

data has model minutes ago, been we're data quality those think improvements with I we attributing customers I performance. have incremental their few releases, to a great working mentioned absolutely, proven to new who, that their So their for quality and And are to customers. lead

there's outcomes do. science data data that are we're there then thrilled LLMs, we're we're there's for work and and to great producing that that and outcomes of growth.In those customers, So There's terms those compute, seeing leading we are engineering expansions critical X great ingredients. our

them -- and strategic kind are of models of that that and they're models investments in risk that emphasis Given building. kind making to they these the companies that will they're and these result accuracy they're on to critical the have that placing developments, it's that the the the of the the data taking

we we we're future and think is great very doing work, So well We're positioned. bright. very think the

Timothy Clarkson

Right.

you one how just of on and obviously, going to this and comment future? then you Can project as wasn't with an historically was into it model revenues. ended you I challenges, kind think mean idea, is there business extension Innodata durability a to do of durable a of be the the new that Now

Jack Abuhoff

We these in We several extremely our it's see all to still opportunities. innings customers the those. developments multiple early more multimodal landing able same and we're time, as We're they of we take able the specificity.All investments. to early such of place. I if to their to at And of them, executing on going companies, to we're opportunity. growth with talk tasks want lots to I are mentioned early So think things that taking of multiyear the languages, still. think about our early for are models starting a be enterprise task be of complex that that domain we're days result getting and in partnerships more drive these multiyear the models, and now about them we're reasoning believe we're of into

opportunity in us. So we of front think there's a growth of amount tremendous

Timothy Clarkson

Right.

expenses if you And $XX days days, were XX% tax. old in you don't call XX%, peaking projects, on but seems once XX% have like good potential on with the it netting sales XX% like you to run back it seems a to I -- out, certainly old after it million. to you've margins pretax separating I onetime those get these the investments, remember solid there's guys when rate. a know historically, got the doing guys Now but you're

Jack Abuhoff

I what We we're opportunity that think there. yes, think pursuing. that's is the So certainly

it's think also important. We

growth, we truly in and very mentioned, our that. about to front have that opportunity we're create a of us be company, investing excited great I to in believe we As an

Timothy Clarkson

on then, tied that of be business? the would right, part investments into these AI Now directly

Jack Abuhoff

Exactly right.

Timothy Clarkson

luck. future. good Well, an Great. exciting Right. It's

Operator

Dana The question from be next Buska will from Feltl.

Dana Buska

you. Congratulations of questions on a wonderful quarter. just I a for couple have

are more about for look a little talk bit demand the models? a on little train enterprise Can talked to training bit their enterprises you seeing You you that? more own data. about And

Jack Abuhoff

so which So the models more down, comes own their I we cost inevitable. we're going is of enterprises see doing believe think to training when

early part, things most -- into that enterprises data and piloted the that become improve the of accomplish positioned be POCs in context think days feeds we're is and in going going now of possible, of doing with we're help do today. even there implementations. with lot help into see very to development past are enterprises opportunities and RAG achieved have into are now those what them. very We're Again, now data that. use be to to models a integrations that development, the we being kinds and of can to those or windows looking own to For things to POCs, of results well quality getting that that, move implementation.We their models is with context what And especially order large terms within from as

Dana Buska

Okay. of come you how seem read you speed may safety. Could about trust Great. talk threat? how are up, and about your fine-tuning consider tools to that do across lot I doing a little In And to affect business? and a to bit coming lines, a those companies software that articles up with be and or

Jack Abuhoff

world consider technology that No. will at an be in and have. more POC have has opportunities to see We we opportunity. success moving the the acceleration The more the that and development, we we the from things into

we a We it's about help that. around data-driven companies among the along that safety. data-driven and trust believe a companies we're are capabilities initiative We're Again, and in developing who taking approach that that is path. excited will very

Dana Buska

looking at Excellent. a hiring to type talk a bit That new little hiring you're people? you're Could positions costs, you you bit talk what recruiting looking exciting. hire? the sounds And of new when about Okay. these about could little

Jack Abuhoff

them to hiring people people that these domain kinds us language very the models. are experts, language. we're we So backgrounds linguists, that enable data require sets they're that hiring train have to ability of hiring they're nuance with build to of the help A appreciate and lot who we're they're the people broadly. experts, We're

And we about it's we that to As used do the better is models them. models, think program performing fine-tuning language around of the these nuances can languages, are. more those the the

Dana Buska

for Okay. targeting last growth a one it about that year? that Agility. question then you a that XX% Is are is about growth It you think like this quarter. that Great. sustainable? rate is rate looked And Or grew

Jack Abuhoff

if was about it think I'm I mistaken. So XX%, not

Aneesh Pendharkar

was Yes. Dana it XX%.

Jack Abuhoff

XX%. Yes.

a opportunities some probably relative do what more on allocation we more large in very lot to. be tech at service certainly with it, I aspire companies. would now, that's that think could something But we XX% point. focused would considering the right with to So have we to we're XX% is reinvestment the We and capital

Dana Buska

very sounds exciting. It Excellent. Okay.

Operator

from The coming Galloway from Bruce next Galloway is Capital. question

Bruce Galloway

Jack, getting you're lot traction. It seems a congratulations. of like

to contracts the new size the Just these of of quantify and some additions.

mentioned You that year to I for that million one you years. doing hyperscaler $XX X in $XX.X it on announced $XX guess that a that the was the April. million add extended was million the

to? which contracts, could base from of scale can of on million So these and customer. that they the much XX% over other quantification they last add-on rate on some how million a you alone year, is just growth big $XX from a million, give one that's how some $XX $XX.X Can us be

Jack Abuhoff

Sure.

believe of have you were is would them will. very I we we're So they appear another big know brought quarter, it to believe mentioned, certainly these that X referred technologies. significant players now significantly be eager And we X to big very which in as us. well with this But spending work as to XX% with as customers reasonably target. scale are very that portfolio quarter, it don't all of when And still the we and positioned.We I to as we terms doubles we're them. on we all a when the of working what adding that numbers in up and tech which possible, adding conservative we start as target, that what's And growth are look last we one? believe that all up you the company provided happen. this believe Will companies. across an significantly if wonderful be become That with to opportunity --

flight. order of right in view lots conservative opportunity a business support We're taking our in guidance, in growth of it's pipeline our to a have now. very exciting time We our and

Bruce Galloway

a at modest So coming personnel all the mean, current huge with do CapEx you have installed amount infrastructure, for the the and mean I contracts looking these you $X seems $X have upside I that very your and in through. your personnel million your base, adding capability? in and scalability million

Jack Abuhoff

that that revenue. cost costs cost we So XX% costs should again, I'm and that to you development. the perspective, assume headcount drive were would that not goods. of probably to But area, look in investments we're be I to the an from and investments as marketing with I And margin. associated a in other as goods about growth, product for services the incurring of the near-term opportunity it get adding required of that investments would adjusted think that at the producing contracts the gross XX% is to in costs about numbers, including the making in the sales, spoke recruiting those

Bruce Galloway

from as And combination that Okay. Is cash. cash you Or thereof? cash it million like is the in Or far receivables? generation, pure also, from $X.X mostly generated as operations? a

Jack Abuhoff

include I operations cash Yes. that mean, from receivables. it's

collect. as So then we build, we

Operator

Q&A now back the Abuhoff closing Thank remarks. that will today's for hand you. And conclude session. Jack to I does call

Jack Abuhoff

you. Thank

raised hope revenue believe additional XXXX of at a of win which both X Longer we also with excited increased we contribute QX. So quarter. XXXX customer And the expected substantial work on to in win, substantial that year-over-year signing off we have this super grew by couple energized least addition big growth are we next annualized base, another revenue for sequential market expected worth trajectory. We we're from we tech we announced expect term, exciting customers, million I of start is customer XXXX. by the the revenue.We and continue will overall, yes, an to same we of our expected year-over-year weeks organic-only underway. growth. revenue. XX% revenue enterprise my the demand growth.Today, that's of $XX ago, We yet we to million approximately team from about what to accomplish XXXX. $XX.X announced And will We're top of believe XX% accelerate our for our a from guidance approximately announced annualized in tremendously revenue valued in big XX% increase customers -- continued to will customer This growth one at tech or anticipate

creating focused are long-term attending on to forward next We be laser looking for today's shareholder We'll value.And our thank call. all call. you

Operator

today's This Thank conference. you. conclude does

your may for at Thank you your this You time. disconnect lines participation.

Innodata (INOD) 7 May 242024 Q1 Earnings call transcript