Perspectives

Building the Future of AI, Together

November 29, 2023

Despite boom AI building blocks necessary pre-train fine tune leverage new foundational models shockingly limited enterprise wants use proprietary model software applications need provision GPUs clean tag data structure database run algorithms route however common remains inflexible often prohibitively expensive better answer future enterprises want sophisticated approach leveraging tens hundreds different concert Companies scratch open source closed OpenAI's GPT-4 order realize vision best-in-class tools create platform orchestrate multiple work together Today excited announce investment Together.ai provide allows shapes sizes open-source host inference across saw frontrunners emerge early evolution SaaS Emergence see patterns repeating leading edge B2B thrilled part journey breakout company provides critical infrastructure develop innovative that'll precisely affordably greater control safety providing access library datasets vital make development easier ready pushed production API hosting now AI-powered solutions transparent flexible cost-effective ever team helm including veteran entrepreneurs well-regarded academics deeply technical long forefront research brainchild Chris associate professor Stanford Lab researches applies foundations next generation systems developed idea partnership Percy Liang Computer Science director Center Foundation founded serve thought partners CEO Vipul Ved Prakash previously sold Topsy Apple 2013 CTO Ce Zhang former Ph.D candidate worked Re rounded Chief Scientist Tri Dao whose seminal FlashAttention v2 key technological advantage customers unique confluence academia entrepreneurship yields exceptional founder-market fit alignment perspectives particularly belief variety ranging Together's enables rapidly cost effectively build using cutting training optimizations helps achieve core piece intellectual property without reliance third parties OpenAI conviction technology give flexibility fast unlock value first layer reinforced beliefs take full-stack burgeoning industry look forward seeing products eager support

Where to Build GenAI Solutions in Healthcare

November 9, 2023

crossroads comes healthcare U.S cost delivering care country escalated degree seen time life expectancy fallen behind developed nations Something change respect better health outcomes lower said potentially good news horizon Generative AI enabling creation new solutions potential improve Founders responding developing offerings already beginning make impact parts industry importantly per recent study Bain Co technology buyers strategies starting purchase GenAI-powered Emergence always kept close eye intersection B2B software identified promise Doximity Veeva early journeys now seeing number exciting companies created using GenAI tackle biggest problems belief generational leverage keep pace tapped network experts making tech decisions impacted across providers payers science include leaders institutions Mayo Clinic Blue Cross Shield Benchling startups League Assort shared best opportunities emerge excited share insights framework think areas target building applications rooted research designed deciding want build heard 25 expert conversations three factors repeatedly came likely guide decision-making unlock interested solve practical Decision makers deliver order magnitude improvement existing Urgency customer pain budget vitamins tools genuine painkillers feel costs high places quality needs major Perceived risk risky little tolerance mistakes inaccuracies slow adoption perception exist degrees nothing binary ripest ones addressing meaningful important consider view offering However significant believe willing experiment provided selling acknowledge working ways mitigate possible mind present following assessing greatest opportunity mentioned earlier provides significantly solution previously existed addresses matter lay gives sense near-term note decided treat post lens layer top ideas plot graph dissuade anyone high-risk yield highest lucrative offer require careful planning terms product go-to-market efforts drive bit detail provider payer segments clear overburdened Doctors nurse practitioners physicians assistants nurses patients day small slivers administrative burden associated remaining compliance managing billing creates mountains back-office work result higher burning increasing rate spoke systems ability delivery Note-taking benefit scribes alongside patient visits take diligent notes ease documentation shoulder expensive makes human-quality first thereby cost-effective Prior authorization approval process go ensure qualify coverage thorny sounds requires 5 hours physician every week automate claim management Medical complex extremely time-consuming right evaluating clinical applying proper codes submitting claims format insurer handled quickly accurately incredibly run insurance company tracks processes millions month powerful distinguish crowded market operational efficiency side businesses Benefits education Educating huge center 80 calls insurers receive fall bucket custom personal nature inquiries hard scale non-human break questions match people high-quality answers issue see develop sophisticated streamline hits-driven business finding next life-changing drug getting regulatory bringing average takes 12 years 2.3 billion bring 15 since last year consistent pressure prices rise trial recruitment Testing demands sample group specific attributes symptoms experiences find researchers need delve unstructured data collected thousands sources ideal task identify patterns population speed read records point pharmaceutical directly trials optimization 90 fail reasons ranging poor design unrealistic deadlines used simulate debug start monitor going reporting great anything turning bunch disparate prose exactly needed report results crucial clock patents starts immediately 20-year commonly halved passes dollars lost generate FDA-ready reports faster outsized bottom lines Marketing campaigns Extensive legal MLR reviews prerequisite promoting critical help create compliant materials messages saving nontrivial profits watch-outs takeaways wildly enthusiastic remiss mention caveats commoditization Given addressed off-the-shelf open source certain competing solving become commoditized Outsourced competition jobs done listed currently customers human-powered services perfect often win cheaper effective options EHRs race Heavy hitters electronic Epic Oracle Cerner sleeping invest features broad 63 combined mindshare Beating creating superior products None challenges deter GenAI-enabled massive unprecedented momentum buy-in among prospects capable move fast overt demonstrable ROI hope frameworks useful whether determining considering set founder love conversation reach kevin emcap.com kyle

Demoing Value with Saleo

October 5, 2023

trend hyper-personalization longer solely associated consumer world now becoming new standard B2B consumers come expect tailored recommendations seamless user experiences content aligns precisely preferences buyers grown accustomed personalized engagement field nail sales demos Enter newest investment Saleo first live demo experience platform Today happy share Emergence led 13M Series round takes lackluster costly status quo first-ever AI-enabled solution creating vivid product Software often time opportunity visualize teams use process friction pre-sales prospect sides Setting time-intensive professionals need rope data science team designers developers hours work render within resulting subpar salespeople prospective customers break nothing update prospects hard truly visualizing value evaluating Bad cost deals addresses critical void market enables companies showcase actual native products dynamic realistic automatically populates changes real competitors offer static image solutions dummy leverages proprietary model empower demonstrate true nature showcasing immersing potential combines AI's computational power human expertise revolutionize buying making vastly simple efficient sellers impact clear experienced 50 shorter cycles 2x increase NPS 74 win rate ROI numbers seldom see best tech Anyone worked bought business pain Chief among co-founders Justin McDonald Daniel Hellerman Alexey Svetliakov Long-time veterans marketing technology SaaS team's combined vision form partnership embodies essence Emergence's Focus Drive Conviction principle aim embody seek founders rarely founder-problem fit catalyst change landscape look forward partnering journey create better ultimately benefiting businesses alike

Meet the New AI Infrastructure Powering the Next Gen of AI Applications

October 4, 2023

investing AI-first software decade Emergence exemplified early investment Chorus.ai sold 2021 575M companies built AI tech stacks scratch Today powerful generative foundational models made publicly available new adopting existing workflows applications faster ever including several portfolio Doximity Ironclad Saleo experimenting LLMs across numerous facets operations deploying copilots using power key internal customer interactions enterprises become innovative sophisticated use realizing want ownership control rely solely closed OpenAI's GPT-4 black boxes defensibility belief leverage tens hundreds simultaneously proprietary pre-train open source fine-tuned specific needs facilitate surge enterprise adoption need infrastructure stack focused speed cost-effectiveness transparency security data IP Building foundation build AI-powered access three incredibly limited knowledge experience required fine-tune unique products Right now simply researchers necessary breadth depth model-building becoming fluent everyday industry develops gap supply demand Compute GPUs train provide availability due capacity constraints GPU manufacturers problem forever production scales time increases amount critical challenge overall solve order keep rising Tooling enables transform raw useful tune model clean tag structure database run algorithms pre-trained value host inference queries result application workflow lifecycle siloed mostly manual step elements essential developing however bottlenecks fraction truly capable leveraging becomes widely matures believe incorporate Ultimately winners created layer come multiple flavors Whether apps replacing human-powered services creating sit alongside business users brand gen core excited invest Cambrian explosion emerging unleash enable concert choose different types trade-offs similar ideal non-customized cases cost-effective rapid experimentation limitations industry-specific augmented RAG techniques better suit company's reduce vendor lock-in future costs reserved high-value highly customized expensive time-consuming develop offer competitive edge intellectual property outlined succeed world ones best match every company orchestrate end announce heart directly addresses Together.ai platform provides everything open-source read area thoughts love hear please reach joe emcap.com

Your Success with Generative AI May Come Down to These UX Decisions

August 9, 2023

long-time customer support leader know recently went shopping generative AI solution processing tickets shared framework using evaluate options came three dimensions Cost Performance UX mostly equal given underlying tech apps different companies better job building capabilities powerful delivering value result Hearing curious making choices integrate implications decisions go-to-market success end users serve clear competitive likely enduring field ones precisely tailor shows products match user needs expectations every step growing commoditization believe thoughtful major determinant behind wins loses generation AI-enabled applications said start digging ton questions without obvious answers pop including chat interface utilized teach people prompt adding friction co-pilot features incorporated best human machine goal piece give handle good actually looks talked dozen product design leaders Notion Guru Ironclad Retool Maze leading edge found central tradeoff lots hard-won advice Trading flexibility ease use SaaS always static asked point click goals LLMs break wide open Chat-based interfaces particular massive problems solve time models adds complexity confusion co-pilots embedded offer constraints trade-off tend useful narrowly defined tasks designers spoke core question clarity user's Whenever less practice opt flexible general choose maximizes learning context move toward constrained make navigation intuitive predictable creating all-important sense common mistake picking single type feeling stick implementation take multiple shapes learn moves along journey evolve around become clearer narrow guide folks happy path doubt progressively add mind dive deeper types proving effective depending course tradeoffs right greeter Natural language incredible innovation others excels conversational requiring visualization confusing navigate via blank canvas instances ideal welcome wagon help direct parts achieve wants words guided direction complex find learns gets onboarding presents almost infinite paths entrances granular destinations strengths deliver experiences feel personalized meeting However B2B software important hand opinionated want go Otherwise lost simply taking long need practices incorporating process great first impression full variety coming getting person reduce dropped point-and-click tool something heard number talk excitement experimenting early days terms live deployments Collect data Telling robot marketer fun choosing dropdown menu depth breadth humans share chatbot amazing glean metadata based ask say able drop optimal portion helpful feedback refine development addition searching information provides knowledge management utilizes employee Little needed trained company-validated cards improvement already established search box task Remember directs greet road next Specialized usher experience form designed interact specific topic task-specific bot offers bounds opportunity Think points walks seat constraining topics discussion drawing tailored dataset aid strategies seen limited Embed goal-specific Monty built G2 helps identify asks set interactive vendors embeds elements sharpen arrives final vendor list links G2's purpose-built evaluation finish releasing within testing dynamically chats prototype AI-moderated research Communicate chatbot's limits Let focused limit frustration calibrate gives effectively personality giving chatbots personalities come workflows require in-flow UX's visual descriptions well-suited importantly cases reactive instead proactive means cognitive burden quite high chance hand-holding coach Enter trades-off increase workflow enabling suggestions integrated approach allows capture tying actions business outcomes recommendations takes perform perspective improve serves across Emergence call Coaching Network investing since 2015 starting Chorus sales conversations Textio HR writing grown increasingly popular now used coding Github contract negotiation hard poorly lose trust ignored maligned Microsoft Clippy overreaching run risk putting autopilot allow AI-generated errors propagate systems propensity hallucinate active participation necessary ensure avoid former closely monitoring confidence thresholds made fewer slowly expanding gather usage target Acceptance Rates percentage taken north 75 trusted relationship intern manager produce draft ultimately responsible outcome work thought worker accelerate accountability must remain key application developing encourages contribute actively rather passively accuracy provide creativity mutations model everyone else operating incentivizing engagement Create assist legal drafting requires double confirm particularly sensitive experimented lawyers input Bar hints highlights phrases change e.g attracting diverse applicants Periodically Textio's highlight phrase suggestion ideas Relatedly TopHat ed-tech company students Align incentives consider compensating additions drive impact example rep modifies drives higher deal close pay commission deals explore done nice surfacing entire possibilities write blog post author poem shorten summarize document translate text bevy languages arrive overnight panel utilizing helped Running discovered translation option turning rough notes helping day determine prioritize roadmap calls faster ever pointing new methods surfaced top 10 pursue perhaps clearly message Button toggle similar Notion's sound professional exactly Unsurprisingly exist spectrum inflexible producing depends entirely surface tips menus added abilities existing toolbar pithy keyboard shortcut functionality puts pathways scrolling makes unbounded though opportunities benefit enhancement heads proactively ability tooltip pop-up muscle memory capability dopamine boosted prune track critically evaluating stay bogged bloat often conversation rapidly evolving Countless answered build maximize AI's tendency hoping continue ongoing builders ecosystem learnings love hear deploying UXs resonated thus far Reach jake emcap.com Thank wanted acknowledge Mike Wheeler Josh Abrams Rick Nucci Kieran Snyder Jo Widawski Jamie Cuffe Linus Lee James Cham Bloomberg Beta Justin Cranshaw Maestro

How to Price AI In SaaS

June 6, 2023

2019 wrote post companies price AI-enabled software focused SaaS developing AI highlighted pricing considerations work improve models Since meteoric rise third party foundational model providers OpenAI MosaicML Service vendors enabled player integrate powerful application created mad dash sprinkle pixie dust across ecosystem seen among countless newly minted startups established public Hubspot Doximity proliferation technology raises questions including deploy safely win incumbents existing distribution important area discussed priced lay working framework think space evolving rapidly update thinking future posts Start understanding differentiated value features creating definition accessible every provider effect commodity integrated product first principles feature create integrating flow broader platform saving user time leave go underlying ChatGPT etc Keeping context unlock see thoughts Deep Collaboration However honest actually today getting flood initial tire kicks curious users seeing meaningful sustained adoption retention creation ask offerings majority creates garnered going directly try make significant margin Reselling sustainable value-creation strategy differentiation though topic another able charge making current valuable perhaps stickier used drive upsell higher tiers result increased net dollar leverage thin wrapper around third-party build point consider value-extractive approach state early days launch priorities learn add end starting offering free leaders chatted behind strategies envision monetizing said already begun freemium typically gated usage volume gets full taste feature's power paywall use prioritize simplicity helps GitHub CoPilot NotionAI strong per seat basis aforementioned incremental fee top core whose entire providing services Jasper Copy.ai deliver copywriting prices likely cost goods sold COGS given bills open question pass along customers case provide truly costs healthy commoditized thus lower gross profiles traditional 70-80 players internal operations e.g automating major parts sales marketing recover loss operating expenses offered level token compute assume better inference continue decline near term stagnation GPU supply constraints portfolio summarization editing text generation breaking bank gating paid reserve right throttle certain extreme intentionally vague fair policies mitigation optimizing task trading performance example leveraging GPT-4 latency despite tremendous GPT-3.5 Turbo sufficient cases model-specific points evolve quickly idea choosing job become dominant take delivering open-source particular dynamic favorable Driving towards Ultimately goal simple integration using capture proprietary data specific done JTBD want track correlate suggestions surfaced suggestion business outcome tune JTBD-specific obtained generic Emergence call Coaching Network Ironclad good provides digital contracting draft negotiate sign contracts recently allows redline started OpenAI's now gathering model's critically outcomes redlines faster contract close timelines insights others accordingly move toward nirvana suspect away limit cannibalize seats less necessary aligned industry share interim solving learning iteration evolves blinding speed Feel reach jake emcap.com building jam ideas earlier version article featured TechCrunch+

The Next-Generation CPQ Solution: Our Investment in Logik.io

May 25, 2023

investing enterprise software Emergence 20 years privileged witness remarkable transformation markets working alongside founders evolving platforms shift traditional on-premise cloud paradigms written extensively Deep Collaboration AI Coaching Networks last decades witnessed types long-standing enduring CRM since first investment Salesforce new rising progressively replace inefficient pen paper workflows prevailing assumption hard innovate within former solved already question Eric Yuan asked hundreds times started Zoom several large established players video communications market around forever problem better know truly leading technical team largest category decade turned knowledge existing products meant knew product offer 10x experience end users short communication solving shy away problems eager always find next iteration make Enter CPQ pioneers configure-price-quote emerged 2000 Chris Shutts Godard Abel decided launch company BigMachines applications quickly cut complex selling challenges giving companies ability streamline service configurable slight improvements configuration pricing quoting guided order management strategies revenue leaders nascent largely contributed democratization CPQ's tight relationship central systems ERP developed strong bonds vendors categories Oracle etc eventually acquired 2013 consolidated Oracle's suite tools significant went lead following eight delivering immense value customers began needed 2014 SteelBrick proud investor building cloud-first natively top ecosystem later necessary broader interact delight seen successive creation coming age solutions offered shortcomings ultimately want Acing today's explains excited today announce led Logik.io's 16 million Series partnering second time co-founders transform B2B model convoluted expensive consumerized low-touch efficient Knowing power together understanding exactly delighted work nine SteelBrick's built Logik.io perfectly happy unhappy detail pain points aligned happening direct sales anymore across e-commerce omni-channel use cases observe iterate front finding immediate product-market fit swiftly signing six figure contracts need Logik's solution Navigating back hand certainly accelerated Godard's timelines thrilled continue partnership Logik come days embodies core values Win Big Long Run

Welcome AudiencePlus: The Owned Media Software for Marketers

May 11, 2023

Emergence delighted announce led AudiencePlus 5 million financing round company begins transform marketing automation industry thrilled partner Anthony Kennada Tyler Sparks believe represent perfect founder-market-fit reinvent B2B companies go market fuel growth first-of-its-kind software platform enables marketers use owned media build sustainable pipeline Imagine relationship audience launch beautiful branded property website without writing single line code distribution channel content beyond views listens understand actual engagement individual account level talk founder-problem fit Anthony's deep domain expertise exactly look known 10 years teams Gainsight Front Hopin building solutions in-house working external agencies expensive time consuming engagements knows overlooked products create fill gaps story reminds Eric Yuan's Zoom worked Webex began thinking idea told web conference space saturated filled large incumbents enough opportunity first-hand experience problems conferencing knew customers existing making happy determined innovation needed today see Zoom's massive success parallel excited jump entrepreneurship side bring best every developed proprietary call Graph collects context topics formats contributors job title behavioral data outcomes impact order train machine learning model using unique four key components modern CMS end real-time analytics customer backend manage first party workflow Similar built different era Now incumbent big organizations longer run founders know pain points fallen behind addressing needs technological particularly AI rewrite entire codebase provide value initial cases include recommended e.g Netflix TikTok generative metadata composing models GPT unlock change users able integrate CRM revenue take great delivered Crossbeam Five9 Lavender Nextiva Zuora make available everyone else already provided creating audiences followers become continue grow Hear team announcement funding TechCrunch

Specialized Software Will Win in the Age of Generative AI

May 4, 2023

Lessons Investor Center 25 Years Software Shifts inflection point buzz around generative AI founders investors scrambling find promising applications technology Q1 2023 alone 1.7 billion invested 46 companies startups continue proliferate money continues flood space want take broader look back learned previous platform past 20 Emergence witnessed trends hype cycles lasted see similarities decades seeing unfold today advent horizontal cloud early-2000s promise industry-specific late-2000s evident specialized builds tremendous value shift early started building broad solutions matured start build specific industries targeted jobs-to-be-done JTBD think applies today's cycle piece surface throughlines apply context On-premise 2000s democratized first time ever web browser instantaneously use run portion business personal life good fortune working Marc Benioff launching Salesforce.com co-founded company called Service became origin Salesforce saw days ability bring Amazon-style simplicity every user allowed easy adoption fastest-growing era including Workday ServiceNow created replaced former client-server leaders revolutionized new model now become standard unique disrupted long-standing distribution channels value-added resellers VARS distributors Instead leveraged product-led strategy rapid impact market players adapt fast enough eventually faded away small number incumbents adopted prospered specialization within dozens more-targeted successes ZoomInfo SalesLoft Chorus focused jobs sales stack picking left products tailored better serve users providers Industry norm second major transition leader Peter Gassner Veeva Systems coined emerging field 2008 leading institutional investment chairman since IPO power data important key hallmark deep create customer entries behavioral analysis potential focus makes far likely turn meaningful insights customers type impactful needs single comes layering cake Whereas broad-based across therefore less-actionable deepened datasets stacking multiple functional areas root perfect example sciences developing 30+ solve industry's challenges able quickly using less capital cousins reached 2B revenue 3M Mass customization vertical slice purpose-built solution meets customer's success words works settle believe step came began narrowed later industry-focused next age proprietary rise popularity trying answer question consumer possibly ask settings similar AI's powerful tightly generic models trained large swaths narrow cases required provide B2B organizations tighter problem valuable face significant differences eras starting difference underlying closely aligned Microsoft Google Meta others largest tech different etc built ground big owning domains APIs open-source easier anyone adopt paths ways forward lasting defensible understand team's core strength choose options technical capabilities hunt function benefits domain expertise relationships form team add knowledge Given accessibility strategies lead opportunities lesson remains true wiser product diving fits landscape founder 20+ confident successful learnings stay something please hesitate reach gordon emcap.com