Perspectives

Pace Raises $46M Series B to Scale AI-Native Insurance Operations

May 28, 2026

Last year 60 world's losses went uninsured 9 trillion problem considering value risk involved insurers bad pricing affording service especially smaller accounts another industry runs faxes legacy portals phone calls insurance expect AI break result software companies keep distance Pace company raised 46 million Series Founder CEO Jamie Cuffe direct line industry's biggest pain points grew around father's work reinsurance brokerage found Retool platform enabled build internal business without writing scratch Now mix fluency product instinct showing Pace's traction since launching agents completed 250 000 critical workflows growing 3x every quarter including Prudential Newfront Palomar alone automating thousands hours policy servicing issuance Behind numbers simple design choice end-to-end submission intake claims handling data entry Rather wait modernize systems navigate already exist clicking apps reasoning across documents Lessons front lines sharpening think AI-native services Domain credibility opens doors cold outreach never Cuffe's early conversations landed background gave instant room turned deployments became springboard marquee carriers followed version economists call Jevons Paradox playing handles quality assurance original observation 19th century economist William Stanley efficient steam engines reduce coal consumption increased cheaper power created new uses worth pursuing operations pattern Make dramatically faster buyers use Traditional BPOs QA sliver human review expensive effectively constraint made sampling necessary gone Fewer ten people handle estimates take close thousand outsourcing model Cost reduction headline real story total addressable market stops competing scale Underpinning success advantage shared become capable underlying LLMs improve pointed single release earlier took rate navigating interfaces 30 95 proud back team structural incidental costs stop reason coverage gets insured

The Deployment Era Is Here

May 11, 2026

past several years defining question enterprise AI whose model wins gone away capable models now available across closed open weight choosing increasingly routing right task cost matters enormously longer separates leaders new taken place best job deploying technology throughout company internally operate externally serve customers comes execution winners ones taking frontier capability embedding workflows decisions actually run business OpenAI launching Deployment DeployCo see shift bottleneck moved gap capital last decade went building look enterprises use stubborn pattern emerges McKinsey found 88 organizations least function fewer 40 scaled beyond pilot Access nearly universal Real rare organizational problem technical Running easy Redesigning core workflow around connecting live data processes getting frontline teams rely different kind work takes people understand organization stay inside long enough solve built exactly Forward Deployed Engineers embed working alongside operating find creates leverage rebuild ship production systems neither software delivery traditional consulting vendors hand implementation Consultants recommendation DeployCo's FDEs system muscle put experienced builders field day agreed acquire Tomoro applied engineering firm roughly 150 already done demanding environments including deployments Tesco Virgin Atlantic Supercell partnering conviction category call AI-native services AINS idea straightforward transform firms generation companies deliver expertise required large expensive humans supplying judgment relationship written mapped early space kept refining thesis market matured largest ambitious test seen Critically designed deploy applying helps clients implement operations works scale resets expectations Emergence invested founding partner wanted close learn team push anyone update playbook patterns emerge expect partnership sharpen thinking considerably consortium behind worth noting launched 19 partners TPG Bain Brookfield Goldman Sachs SoftBank 4 billion initial investment private equity sponsors alone back thousands portfolio structured path serious bring change-management depth provide deliberately constructed coalition reflects seriously operational side transition advantage compounds keep improving expanding possible starting compound somewhere else movers refined institutional knowledge stuck mode playing catch-up good bet solved specifically moment thrilled journey

Should Your SaaS Company Become a Services Business?

May 5, 2026

Sell Outcome Seat People never wanted buy software always means end logical conclusion taken seriously job done decades best approximation made faster reduced errors lowered costs required someone sit use tool customer bought hired person hoped combination produced actually AI changes equation first time technology company credibly deliver whole raises provocative question every SaaS leader asking now full scope work become services business want precise describing easy confusion know watched happen real weeks ago raised idea board meeting portfolio companies scaled room confused stares immediate reaction mean hire professional fact instinctive interpretation tells deeply mental model embedded becoming AI-Native touch responsibility delivering result entirely credit spent since exploring partnered space actively evaluating pivot turned genuine curiosity logic clicked Three Reasons Make Leap 1 buyer disappearing uncomfortable reality facing getting laid Block recently cut significant portion workforce dozens across tech beyond makes individual workers productive need fewer used log day champion renewal years Corporate layoffs paradoxically tailwind headcount stock bump often infrastructure place absorb lost capacity needs Outsourcing guarantees becomes attractive rehiring especially vendor cheaper consistent quality businesses dynamic headwind seats licenses lower NRR shrinking addressable market macro trend threatens revenue fuel built underlying product addressing larger TAM 2 defense Claude asked version Anthropic friend foe honest answer Foundation models better extraordinary rate Features took engineering teams months build approximated well-prompted GPT minutes primarily UI top reasoning blast radius guarantee outcomes draft contract take error deal analyze financial data accuracy accounting auditor write marketing copy campaign performance taking occupy position structurally unlikely pursue OpenAI Google building general-purpose engines going outsourced department managed legal team liability domain specificity relationship assurance layer value add difference selling competes sits exposed defensible 3 ROI Die hard establish causality era CFO's mind ever AINS provides path away constantly justify existence Switch working variety founders transition learned expect competitive set gets dramatically easier competing best-in-class attention RFPs well-funded startups entrenched incumbents products decade service providers slower expensive less tech-savvy though claiming added landscape favor Outcome-based pricing forces good fast dollar tied success fee afford inefficiency sounds terrifying tell liberating automate everything automated focus human effort matters Incumbent passing inefficiencies along clients constraint advantage finally feedback signal whether customers continue pay see actual engagement generates improves next reinforcement loop shifts renewing specific piece produce right founder told single biggest unlock embarrassed embarrassment gift primary user feel workflow gap broken integration clunky UX choice development velocity increases shrinks quarterly NPS surveys daily lived experience entire go-to-market posture change longer technical expects startup pitch operations trusted provider Present trust well-respected practitioner industry AI-pilled interview filter ask candidates explain already tools nothing fit Valid Objections pretend barriers making DNA transplant completely different muscle moving toward something looks firm org design drastic experts delivery sales motion run likely decisions part teammates whose skillsets mindsets aligned new direction shift harder bigger existing extremely difficult let go scale book roadmap demands keep standing classic innovator's dilemma generating today unproven takes courage start shifting resources Gross margins existential converted high-margin low-margin downgrade stop thesis depends lion's share humans handling exceptions relationships margin trajectory important metric sure calculate correctly including salaries inference spend improving quarter staffing Defensibility requires intentional inherently fungible platforms Switching BPO ripping integrated system think deliberately sticky avenues Proprietary assets accumulate fine-tuning valuable archive reference access leave-behind Palantir Foundry gives foothold ends Deep customer's systems record pulling disruptive brand track demonstrated compounds deeper strategies Playbook Apply sense five true persona threat ones fight budget line consolidated buyer's urgent trigger strategic option necessity function serve critical non-core differentiates happily hand Fund administration insurance brokerage compliance reporting cycle management essential functions rather views core identity outside owning Meaningful flows firms category consultants agencies BPOs buying exists creating redirecting study learn weaknesses enter 4 repetitive enough managing supports highly bespoke judgment-intensive AI-leveraged diagnostic simple high-volume standardized low-volume unique 5 genuinely believe produces alternatives honesty Committing avoid move needle hoping figure hook directly stake reputation conversation worth Waiting Risk understand feels ARR net retention investors love walking dive reckless consider alternative curve slow product's functionality replicable pressure automation moat narrowing worst story wait watch decline 120 95 experimenting agents replace try weakness strength exponentially big bet adapt risky deluding thinking relative status quo safe thrive optimizing year comfortable metrics Invitation hear published detailed covers mechanics structure measure updated follow subscribe find LinkedIn

Robotics Isn't One Market. It's a Stack.

May 5, 2026

conversations C-level executives manufacturing logistics companies conversation shifted whether deploy autonomous robots fast industrial labor shortage going away U.S manufacturers roughly 500 000 open jobs every month past six years demand clear long time now first technology beginning show tangible signs improvement robotics market distribution complex LLMs accessed via simple API end customers solutions distinct requirements metrics care see landscape think biggest built Real Urgent case settings always intuitive dangerous repetitive increasingly hard fill energy healthcare shortages acute trucking already tops 80 drivers projected reach 160 2030 total industry employment sits 2022 peak 2.4 workers nearing retirement new entrant 25 Today's Fall Short today confined highly structured tasks palletizing pick-and-place fixed assembly lines task requires painstaking manual programming extensive on-site testing significant integration work standard robotic arm costs minimum hundreds thousands dollars procure Implementation match exceed hardware kills economics highest-volume applications fundamentally statically programmed fail moment environment changes excludes enormous swaths potential use cases represent latent physical world Shift Foundation Models Change Equation different emergence generalizable learn across adapt transfer capabilities without reprogrammed scratch Intelligence investors building layer Early pilots showing deployment times shrinking months days transferring quickly robot types dream state longer science fiction bring turn train briefly works Eventually able hire getting attention rich ecosystem behind ata toon-device compute simulation must come together working Stack map parts infrastructure required make possible become viable result bet single model generalize environments rather requiring task-specific credentialed research teams Progress enable true general purpose watch Generalist Skild AI Field Dyna Genesis Rhoda Data Collection Observability exist scale Multiple approaches competing teleoperation egocentric human video UMI Raw annotation useful training meaningfully less mature exists language vision gap needs close XDOF Encord Mecka Foxglove Neuracore On-Device Inference Running reduces latency cloud dependency critical real-world Memory constraints thermal management unsolved trajectory points toward operate autonomously matures Hailo Axelera DeepX Apex Synthetic let generate experience reducing cost risk fidelity improves becomes meaningful lever accelerating pipeline Applied Intuition Flexion Pays Inspection Maintenance Structural scarcity hazardous wind turbines pipelines power grid utilities pace spend 2 trillion next five Dedicated Boston Dynamics Gecko Aerones DroneDeploy Korial AssetCool Center Operations 7.6 announced capex 2031 creates bottleneck Rack installation cable routing precision-critical making strong fit automation RoboForce Watney Boost Gradient Droyd Warehousing Warehouse median 18.12 hour nationally pricing closing faster categories evenly Unloading kitting remain among largest pools economy Ultra Zipline Locus Pickle Nomagic Gather Tutor Contoro Heavy Construction additional year meet operator workforce ages winning approach far retrofit autonomy onto existing e.g strapping software excavators dozers haul trucks Bedrock Dusty Charge Teleo Monumental Persona TerraFirma AIM Intelligent Machines Cosmic Defense Security Recent conflicts made drones central modern budgets following European Allies Canada spent 574 billion 2025 nearly 20 Counter-drone capability mandatory seeing innovations Anduril Saronic Saildrone Asylon BRINC Splash Surgical facing shortfall surgeons 2036 aging population hospital margin pressure incentives embrace systems pay recurring contracts ROI measurable operating room throughput complication rates Andromeda skilled trades factory production ramps reshoring efforts processes including welding metal forming precision significantly accelerate timelines Path Machina Labs Bots Mind mimic Sunrise Vehicles proof embodied robotaxis complete paid rides per week Adjacent segments middle-mile off-road regulation Waymo Tesla Nuro Aurora Waabi Humble Agriculture Farm chronically scarce input keep climbing growers per-acre opportunity large likely involve retrofits machinery Carbon Beewise Bonsai Orchard Life Sciences Scientific scarcest drug discovery goes bench Lab create closed loop wet verifiable hypothesize robotically execute analyze repeat Lila Opentrons Medra Zeon Home consumer TAM safety concerns solved category mainstream changed price form factor ship approaching high-end appliances improving cover basic home-related 1X Technologies Sunday Matic Weave Prosper Series actually proliferate throughout honest answer layers underdeveloped compound problem makes coming publishing pieces bottlenecks believe matter system integrator enterprise buyers others conviction permanent solvable problems define wave company creation space

The Next Token Supercycle: Opportunities in Long-Horizon Inference

April 15, 2026

Quick Summary Long-horizon tasks becoming dominant AI workload driving huge amounts new token consumption shift unlocking wave infrastructure opportunities see three promising areas Decouple memory compute chips address worsening shortage Build throughput-first inference clouds improve economics compared current providers generally optimized low latency Create generalized context garbage collector prune maintain windows Shifting Paradigm 2023 early 2025 enterprise applications revolved around synchronous chat-centric experiences profile low-latency relatively little persistent late different pattern began emerge apps started focus long-running background requirements forgiving amount input reasoning data often orders magnitude Claude Code platonic product type across application layer example AI-Native Services another major area firm account bulk service delivers outcomes asynchronously follows load asynchronous Matters OLAP Parallel clearest rise great players last decade companies running analytical queries cloud-scale datasets workloads became dramatically data-intensive demanded stack produced several category-defining worth 200B today imperfect comparison share similarly shaped large-scale ones believe necessitate infra bold founders looking era Opportunity 1 Deliver Snowflake Moment Unshackle usually bottleneck due decode heavy nature models perform myriad intermediate steps tool calls various stores MCPs output tokens actions drive length immensely causing increase KV cache ever-ballooning quickly consume available high-quality GPU Without effective pruning compression moderately sized 70 billion parameter model hit wall 100 000 Note 100K 300 pages text Frontier trillion parameters push towards 10 Mythos problem intensify theoretical effects crunch already felt seen surging stock prices manufacturers Now interesting vectors currently explored Framework-level optimizations e.g SGLang vLLM allocate efficiently increasing on-chip local capacity Fractile However truly achieve infinite need essentially analytics 2010s challenge decoupling storage extremely difficult single query small number large database scans requiring network thus making remote viable attention requires access every generated massive utilized simple analogy Imagine copy book word hand desk continuously library town must walk back home unimaginably slow typically required Despite seeing emerging approaches aim loosen locality constraint closer holy grail unlimited networked Notable techniques include transferring optics instead copper wires bandwidth someone successfully externalizes maintaining reasonable unleash fuel generation agents 2 Throughput Cloud Embrace superior Today's neoclouds built speed emerged support first-generation chat-based ultra-low everything rely fleet always-on premium GPUs sacrificing utilization favor responsiveness flip vastly larger creates architecture purpose-built high-throughput less concern theory able meaningfully better building category aware Batching pooling part answer obvious levers chip gains fundamentally capped durable advantage provider extent hardware routing likely source lasting differentiation coordinating heterogeneous prefill compelling technically startups tackle definition education harder markets multimodal Fal winning owning mindshare distinct modality image video abstract shape result risks perceived incremental 3 longer total state want hold always outstrip space window brutal reality retrieval begins degrade long filled reason solutions intelligently compact replace eliminate excessive builds portion work involved whenever talks harness Frankly unclear whether standalone generalizable solution Classic system collection operates deterministic rules remove objects referenced rest using fixed algorithms contrast management judgment involves deciding information relevant summarize losing important signals varies significantly use case fully possible open eager proven wrong Closing Thoughts taking placing unprecedented strain budgets challenges head-on defining thinking love chat live

The AI-Native Services Playbook

March 30, 2026

Spring 2026 Edition Emergence believe AI-Native Services defining business model AI era read playbook idea selling service powered primarily possible 2023 Founders building space figuring foundational elements new go aim diligent students emerging working early pioneers category within outside portfolio document lessons build look quite different successful software company counterintuitive evolving operators considering companies update every six months capture latest AINS share want hear Team Domain expertise must traditional SaaS product credibility important existential customers need ability deliver results established goes long towards trust evaluate reputation perceived person people providing Past performance reliable predictor future Industry experience especially helpful customer-facing members buyer recognize places worked ideally trusted brands legacy providers come co-founders emerge days Mechanical Orchard CEO Rob Mee previously ran Pivotal brought instant enterprise buyers plus access Labs talent network Harper Pace Dakotah Rice Jamie Cuffe grew insurance families giving authentic day Harper's co-founder Tushar Nair engineering depth together served 5 000 businesses 13 Hanover Park Hired senior Standish players fund administration helped legitimize offering built approach ground Crosby Legal Ryan Daniels top rising lawyer leading firm focused practice areas authority unlocks high-quality channels enables rapid staffing maturing Hire leader earlier think delay hiring since customer rarely interacts directly mistake complexity requires strong bridge deployment Without roadmap gets driven whoever shouting loudest rather deliberate strategy productization sales process include someone deep understanding delivered failure mode seen seller intimately familiar delivery overpromises timelines putting bind Involving engineer technical resource ensures fully understood engagement scoped succeed II Product-Market Fit Mirage PMF illusion created revenue growth human labor leverage Beware beast net dollar retention mask lack true enablement Unlike logo prove truly material work high gross margin delivering superior outcomes Otherwise good financed wrong kind capital Real proving scale non-linearly relative costs drive measurable improvements cost quality speed three know Watch warning signs flat declining grows margins expanding honest COGS inference API spend human-in-the-loop belong offload operating expenses absolutely per employee ARR FTE improving simplest test whether pulling weight granularly isolate service-relevant FTEs understand time human-heavy growing linearly base scaling Bespoke significant custom productizing point north star metric needs single number captures actually talk specific effective metrics section sharp ICP extremely precise ideal profile harder unwind client engagements mismatched consume enormous resources distract create requirements transfer broader cases easier productize starting downmarket homogeneity fewer box requests smaller compared Lower ACVs force afford throw Simpler standardized workflows mean edge handle means demonstrate faster focus Main Street daycares manufacturers restaurants Fortune 500 accounts choice illustrates dynamics focusing always correlate larger enterprises developed practices repeatable jobs done take becomes workflow training data oversight Premature breadth fastest paths expand later Strala exemplifies discipline specifically claims processing carriers captives MGAs narrow lets automation III buys implements implementation support function core sell makes around learn operationally part onboarding competency headcount Staff pilots dedicated specialized group runs become experts managing extreme uncertainty pilot roll full project SEAL takes success better integrating steady-state continuity skill sets Sleep customer's office sound crazy commitment follows ensure migration hardest handoff Over-invest level seem absurd It'll pay Sit doers next builders goal rotate learnings flow Law lawyers coders pairs creating real-time feedback used run evals big batches switched give engineers hours allowing tweak system prompts improved meaningfully switch said given enough improve replicate existing processes rebuilding currently exists eng first principles AI-enabled Start board meeting health nuance surfacing problem-solving IV faces fundamental tension paying today entire venture thesis depends tomorrow either resolved destroys Navigate vs urgent tradeoff demands loud immediate Internal platform investments quiet compounding neglecting development responding make competitive market sacrificing distribution another rule thumb started see patterns 80 clients preferences rules exceptions felt comfortable say balance forces best hard element right map tempting find ways prioritize automate fall prey expansion tasks humbling realization techie eliminate role task basis instead handles repetitive ones framing realistic palatable nervous replacing teams Thinking leads automating process-oriented pragmatic Timon Gregg Set tracks impact across vary tie Go-to-Market demo stupid demos showcase value art forgotten default pitch decks understandable provider wields magic internally huge visualize similar startup competing critical hide behind curtain halve cycles adding demoing power effectively Cursor COBOL saw cut half never use seeing action converts skeptics believers far slide deck stop telling prospects Show em Develop partnerships key accelerator incumbents major accelerant offer proprietary datasets crucial corpus small partnership models beyond revenue-share approaches startups exploring equity relationships Others acquiring Prosper AI's Firstsource cycle provides deal years independently trend actively seeking partners capabilities fast Tactical mid-sized incumbent incentivized care top-level influence Deprioritize biggest initially move slowly though relationship feels validating maintain end disintermediated Losing direct flywheel train heaviest-lift able division change improves exactly maintaining matters VI Pricing technology moving toward outcomes-based explore face challenges attribution add using line copilot autopilot remains blurry uniquely positioned definition hires outcome accountable getting problem natural home outcome-based economy structure nature Discrete large-scope well-defined e.g migrating mainframe complex claim price piece clear scope bounded understands Continuous variable ongoing varies handling stream submissions credits-based Credits let normalize sizes tying output credit unit something intuitively practical path norm typically labor-based learning efficiently fine timeline transition matures stabilizes stay cannibalizing increases Embed recurring operate industries multi-year contracts long-term administrator benefits admin last built-in stickiness Palantir customized analyses charging leave-behind VII Defensibility Moats generates byproduct generated advantage predictable MSA letters lead call email policy feeds matching underwriters driving higher conversion rates hallmark great compounds durable uses tools underappreciated form optimally match suited serve style capacity factors Brand powerful moat fundamentally extend professional mattered 4 command premiums despite often mediocre address cold respect borrow respected partnering branded enable switching operations prohibitively difficult record ERP infrastructure layer history operational magnitude integration enormously painful qualitatively simple VIII achieving rate sufficient categories indicators tell lagging confirm showing financials Law's version HURT Review minutes required zero define Per-customer breakdown actions granularity essential identifying stuck labor-intensive honestly bury meaningful compute opex artificially inflates compiling benchmarking analysis plan release summer IX M&A Consider acquisitions eventually acquire rapidly bring timing buildout reason venture-backable place AI-first culture solidified pursuing veterans join conform vice versa Behavior changes mature proven risk turning conventional sheet Comes collapse integrated wants owning stack guarantee structural 5-10x throughput 50 opportunity extends gotten attention vertical high-volume candidate disruption brokerage customs management KYC compliance supply chain BPO alone represents particularly heavy stitch multiple systems generational qualities pick genuine hoping resist seduction measure love alongside thoughts pushback examples host regular workshops founder sessions San Francisco dataset help community shared hub Special thanks CEOs informed Chris Hladczuk Xavi de Gracia Madhavan Ramanujam framework teammates contributions research

Why AI-Native Services, and Why Now

March 23, 2026

AI-Native Service businesses AINS leverage AI deliver services faster better cheaper incumbents customer buys result indifferent performed human hybrid holding vendor accountable single point responsibility built ground around rather assembled acquisition first wrote model April 2024 piece titled Death Deloitte argued companies hollow incumbent conglomerates inside Big 4 alone pull 200B+ year delivering work increasingly commodity age generative inference clearly touched nerve issued pointed response demanding article retitled legal skirmish confirmation fragility world's largest consultancy worried blog post innovation pipeline know moat dried Since opportunity come sharper focus market longer waiting tools results Insurance brokerage rebuilt Harper Tech consulting disrupted Mechanical Orchard Fund administration rethought Hanover Park law reimagined Crosby Healthcare revenue cycle management transformed Prosper founders stakes changed Building company fundamentally different SaaS business possible 2023 great examples learn Palantir useful reference blend software reach elite valuation multiple true playbook startup scratch written real-time SaaSpocalypse Tailwind broader undergoing painful Darwinian transition Advances foundation models raising existential questions whether replaced agents directly correction crisis utility begin seat-based license dying laying people use ARR risk gone corporate layoffs create gap firing doers place replace realize deficit likely prefer outsource vs rehire labor creates meaningful tailwinds Indeed consider pivoting e.g stop selling tool start job done leveraging existing proprietary engine bypass shrinking budget tap massive world original end-users seats owning outcome remain indispensable Action best understand see outcomes legacy providers simply match gross margins make venture-backable Emergence committed definitive authority new era observing shift sidelines codifying high-stakes mechanics scaling defending working shoulder-to-shoulder portfolio community aggregating lessons real time end living resource lead 3T Native tactical guide breaks structural blueprints founder currently trenches leader navigating pivot home base continue document share learnings collectively aggregate standard next decade professional Go

The End of Human Duct Tape: Why We Led Hanover Park’s Series A

March 18, 2026

Fund administration powers trillions dollars global assets Every capital call distribution waterfall calculation LP report audit trace runs botched reconciliation delayed damage relationships took years build stakes enormous infrastructure funds run back office general ledgers built entirely different businesses accounting done across patchwork spreadsheets multiply time new entity gets added roster credentialed accountants spending day work follows rules people talented systems forced use SS&C 20B market cap serving old tells everything size opportunity overdue fundamentally approach Today announcing Emergence led Hanover Park's 27M Series alongside Lux Susa Ventures Rebuilding Stack Scratch companies attacking space taken easier path automation layer top existing sell SaaS tool let figure rest Park something harder ambitious Co-founders Chris Hladczuk Nick Puljic ledger engine investor portal portfolio management full piece data flows AI actually bolting intelligence onto legacy plumbing designing around phrase model requires human duct tape Hundreds holding together processes technology handling replaces purpose-built software keeping expert loop judgment calls matter review system prepares catch edge cases ensure accuracy stake results speak grew 1 billion 15 twelve months completed migration Asylum covering 30+ legal entities decade historical 100 hours industry standard measured Customer Obsession Competitive Advantage relentlessly focused customers success famously sleeps offices successful deployment particularly intense stretch came house napped couch daughter kept watch level tends signal strategy migrates LPs interact auditors pull internal reporting Earning trust kind founder personally guarantee point falls cracks AI-Native Services Playbook service walk away take outcome buys result indifferent whether performed hybrid single responsibility improves accountant serve margins expand faster cheaper compounding dynamic makes venture-backable traditional admin never building conviction past first wrote April 2024 argued core incumbent conglomerates now invested behind thesis multiple industries Mechanical Orchard tech consulting Strala insurance claims processing Harper commercial brokerage Prosper healthcare revenue cycle pattern investments defensible right outcomes held accountable fits anything seen high-volume deeply rule-based consequential enough buyers need someone scales linearly headcount compounds documenting learning update six significant coming soon shaped large part seeing team close collaborators shaping learnings learned amount watching architecture proven deliver pace proud lead round join comes next Welcome family entire

Harper: Building the AI-Native Insurance Brokerage Every Business Deserves

February 25, 2026

Emergence developed deep conviction around call AI-Native Services companies sell software deliver outcomes leveraging AI core faster cheaper better incumbents written model invested behind seen early-stage category achieve remarkable results Today thrilled announce led Harper's Series bringing company's total funding 47M set find company redefine massive commercial insurance distribution market Harper exactly looking Never Boring Part Dakotah Rice first told start TAM slide started uncle's trucking Alabama business fail bad operations financing costs crushed chance talked growing helping dad run nightclub seeing firsthand systems meant protect small businesses consistently failed story captures problem industries gap work actually staggering U.S enormous highly fragmented dominated regional players running manual workflows processes changed decades daycare healthcare facility fleet operator restaurant group trying covered experience painful weeks waiting incomplete submissions brokers stretched thin help consequences abstract Without coverage open logistics onboard drivers construction firm break ground puts supply demand Complex needed real expertise impossible scale changes Enter Founded 2024 brokerage future Rather selling tools existing approach founders explored quickly abandoned broker customer relationship built every workflow proprietary models handle judgment-heavy unstructured slows traditional reading applications routing following underwriters managing quotes deeper unlock matching seasoned know dozen carrier relationships enough route risk effectively synthesize across entire network wholesalers MGAs carriers simultaneously weighing appetite pricing patterns underwriting preferences best fit individual human hours year served thousands America thriving risks brokerages struggle serve efficiently broad describe higher quality process making preferred partner driving strong repeat Compounding Flywheel lead email policy generates data feeds improving conversion rates time hallmark great product generated compounds durable advantage owning full service delivery point solutions ecosystem controls operational make interaction sidesteps classic enterprise sales adoption challenges industry competes directly delivering used reserved largest Founding Team require rare combination domain credibility technical depth CEO brings sharp financial mind honed Goldman Sachs Carlyle Coatue combined family history gives visceral understanding serving ability build trust intensity drive exceptional Tushar Nair CTO former engineer founder AI-enabled Together balance earn relationship-driven ambition relentlessly push toward automation team's execution speaks customers dozens states Ahead become efficient data-driven Fortune 500-level believe kind something generational excited Welcome needs check want matters hiring

Institutional Judgment Is Finance's Last Unstructured Asset: Why We're Backing Rowspace

February 25, 2026

last five years Michael Manapat taken long walks around Noe Valley talking problems worth solving kept circling back pattern seen play across Emergence's portfolio successful founders bring lived experience problem tackling Eric Zoom VP Engineering Webex Peter Veeva run Salesforce's platform win smart inside machine head ML Stripe CTO Notion helped usher company AI era rare understanding take messy proprietary datasets structure actually reason high-stakes environments brainstormed spaces returning investing arena tremendous amount data massive barriers making use going matter finance firm's private history tools clear bar Today thrilled co-lead Rowspace's Series alongside Sequoia 50M total funding seed partnership Drowning Starved Global financial institutions manage 100 trillion assets largest equity firms credit platforms decades deal flow internal memos covenant analyses operating models reviews post-mortems theory dream input layer practice fragmented SharePoint folders CRM systems Snowflake tables email threads legacy file structures analyst-built spreadsheets never meant scale Institutional knowledge distributed people ways difficult unify nearly impossible programmatically generative wave triggered surge vendors promising unlock layering LLMs workflows retrieve information draft summaries answer complex queries documents retrieval reasoning real economic value sits gap pointed inconsistently structured inherit mess Hallucinations increase Edge cases break Conflicting surface without context underwriting monitoring mostly right good enough structural flaw first-generation products skip hardest step transforming something reliably reasoned Valuable market already knows desperate clean standardize process outsourcing Acuity Partners Crisil generate close nine figures annual revenue structuring reconciliation analytical support projected reach 157.5 billion 2030 services driving significant share demand work almost entirely labor report delivered dataset reconciled embedded dissipates static files Rowspace starts different premise transformation service infrastructure foundation unifying firm intelligence top means reconciling inconsistent schemas historical machine-readable formats normalizing mapping specific resolves place search agents path reliable plays distinct powerful PE evaluating new draw assess risks opportunities growth investor allocation decisions act true numbers weeks reconcile matching macro view ensuring compliance loan level managing hundreds billions dollars using exactly chose generic deliver specificity uncompromising accuracy require Public Commoditized Judgment getting fast everywhere Surface-level insight becoming cheap need remains partner hundred deals recognizes patterns others miss multiple cycles breaches noise signals interpret metrics second third-order effects enormous trapped thesis creates advantage reflect reasons conflicts prioritizes precedents becomes auditable memory compound store stores logic behind puts Imagine forgets experienced investor's codified multiplied first-year analyst tap scales instead diluted Built Who've known explored dozens ideas together through-line always refusal shortcut foundational ran applied learning Radar Capital Sigma owning fraud forecasting reliability precision existential later served scaled high-performance teams pushing ahead co-founder Yibo Ling brings complementary perspective two-time CFO synthesized make operational friction team experimenting edges understand decision-making realities markets combination Alfred Lin joining board excited Matters Emergence big enterprise-scale definition trust consequential architecture leaves every traceable exposed early grounded record rather abstract average reset time org chart changes discussing build durable space official Welcome family entire