Quarri
The AI adoption platform for mid-market.
The Problem
SaaS platforms are rigid - so mid-market uses slow, error-filled spreadsheets to bridge the gap.
90%
Of mid-market firms on spreadsheets
94%
Of spreadsheets contain errors
65%
Of analyst time on data gathering
The cost math · Sources
It's not a bug - it's a feature. Rigid SaaS systems force manual workarounds in Excel, manual workarounds are time consuming and introduce errors. Solving the root of the problem requires a risky full stack migration or building a data team at a cost most mid-market companies in low-margin industries can't justify.
SaaS is Rigid and Switching Costs Are High
In mid-market, ERP implementations cost $150–500k with an average of 189% cost overruns (Panorama 2025). Still 30–40% of total IT spend ends up on shadow workarounds (Gartner).
So Businesses Default to Excel
70–90% of firms still rely on spreadsheets across finance and operations (AutoRek 2025, n=500).
But Excel is Slow, Fragile, and Error-Prone
94% of business spreadsheets contain critical errors (Frontiers of Computer Science, 2024, peer-reviewed). Teams spend 65% of their time on data gathering and maintenance, not insight (FP&A Trends 2024, n=379).
The Answer Exists - But It's Prohibitively Expensive
The solution is well-understood: consolidate, clean, and warehouse your data. But that requires $400k+/year before any output and requires constant maintenance. Mid-market simply can't afford it.
Data Team Cost Breakdown
Chart
The Solution
One knowledge layer that reads and writes
Legacy ERPs
Spreadsheets
Databases
PDFs & Scans
External Data
→
Quarri
Read + Write
Single source of
knowledge
→
Automated workflows
Reports & dashboards
NL queries & insights
↩ Write-back into Quarri
Quarri eliminates the need to migrate ERP. No expensive platform transitions, no complex implementations, no larger technology team. Plug the gaps in what you already have — and make it fully scalable.
How it works · Time to live
Mid-market businesses don't need more verticalized SaaS. They just need to be able to harness cutting edge agents.
< 2 weeks to go-live impact
Agentic onboarding skills do the work that traditionally needs a data team - and absorb the cost into the SaaS fee. Most of the bottleneck is client-side data access, not us.
Weeks 1-2
Connect data - ~0.5 day per legacy source. ERPs, databases, external data sets. Absorbed in SaaS fee.
Weeks 3-4
Live dashboards, natural-language queries and external benchmarking. Customer is interacting with their own data. Absorbed in SaaS fee.
Month 2+
Optional · Paid professional services for workflow automations.
✓ Proof point - second customer live in < 2 weeks.
Source: Quarri Business Plan v3 (Pricing Model - setup days per source 0.5, absorbed in SaaS fee).
Agentic infrastructure and white glove results
Quarri uses an in-house AI agent to clean, connect, and contextualise fragmented data, then delivers insights and automation through Anthropic's Claude. No data engineers. No months-long implementations. No platform risk. Instant value.
Data Agent
Automated SaaS - Central contextualised company platform for AI enablement.
78 Tools - Used Autonomously by Claude
Chart
Each tool is an MCP capability that Claude can invoke autonomously. They span extraction, transformation, modelling, analysis, visualisation, metrics, rules, general skills, and admin - covering the full stack traditionally requiring 5+ specialised roles.
Workflow Automation
Professional services - Accelerated automation and reporting deployment for instant ROI.
Deployable Skills
Detail
Teams are trained to use platforms - and we can support with implementations.
  • Natural language interface - non-technical users query data directly via Claude
  • Benchmarking - benchmark pricing, performance, market share etc. against public data sets
  • Data creation and capture - writes back into database based on workflows
  • Systems reconciliation - reconciles systems of records, replacing excel workflows
  • AI-generated dashboards - live, saveable, shareable
  • Deep analysis agent - commercial metrics (Churn, LTV, CAC, MRR, etc.)
  • Document ingestion - ingestion of handwritten documents into structured data
  • Financial reporting - data embedded directly into Excel and PowerPoint
Quarri data agent - messy data to AI ready
Quarri takes in structured data from legacy ERPs, spreadsheets, databases, and external data sets. No data engineers. No months-long implementations. No platform risk. Instant value.
Almost the entire pipeline is agent-managed. The warehouse has read and write abilities - becoming not only storage but also required for data capture and workflow outputs.
The reality today
Quarri moves companies from this...
Messy Excel spreadsheet
SAP legacy system
With Quarri
Automate. Optimise. Analyse.
01 Workflow automations
02 Optimisation
03 Analyse
I showed some of the Financial Op's team too, super impressed. Love the comparisons to CPI, benchmarking to golf clubs in the region for green fees, and the financial controls items on voids/discounts.
CFO
Private Golf & Country Club, North America
What Quarri Automates
Automate recognisable workflows using the power of AI.
Natural language queries
Systems reconciliation
External data ingestion (e.g. weather, CPI)
Forecasting
Automated document ingestion
Reporting & dashboards
Cross-company analysis
Exports into Excel & PowerPoint
Deep analytics
Smart inventory management
How Quarri Works
A cheap deterministic layer beneath probabilistic LLMs.
Probabilistic
LLM · Flexible reasoning
↕
Deterministic
Quarri · Encoded workflows + Structured data
$ COST OF AN EXAMPLE WORKFLOW Human time LLM compute Quarri Data processing capacity 3h human Human only LLM compute 30m human Human + LLM Human + LLM + Quarri · ~2 min 30 MB 40 MB 1 TB+
Cheaper. Faster. Vastly larger data.
Why Now & Market
AI is transforming enterprise. Mid-market is being left behind.
$63.2B
TAM
70%
Struggling to adopt AI
60%
Of AI projects fail - data quality
Market sizing · Competition
This leaves low-margin, operationally heavy PE portfolio companies with the most to gain from AI - but the least capacity to build it themselves. AI adoption isn't a nice-to-have - it's a real-terms EBITDA opportunity.
Market Opportunity
$63.2B
TAM - US + UK + EU + Canada
SaaS Mix > 80.0% · ~86% Gross Margin
$21.2B
SAM - Target Industries
SaaS Mix > 72.1% · ~81% Gross Margin
$3.9B
SOM - Addressable Now
SaaS Mix > 67.6% · ~79% Gross Margin
Bottom-up sizing: US Census Bureau, ONS, Eurostat, Statistics Canada firm counts by size band × Quarri ACV per tier. Filtered to target legacy verticals only. Gross margin: SaaS ~96%, Services ~33% ($600 contractor / $900 charge-out). Source: Quarri Business Plan v3.
Bottom-up market sizing in data room
Why Now
Mid-Market is Falling Behind
Larger companies are 2x more likely to scale AI. Companies under $500m revenue are stuck in experimentation and pilots.
Mid-Market AI Adoption Gap
Chart
Source: McKinsey Global Survey, State of AI 2025 (n=1,993)
C-Suite Demand Signals
60%
Of AI projects fail to move from pilot to production due to poor data quality - Gartner 2025
87%
Say AI Will Be Critical to Operations - Deloitte '25
54%
Prioritising AI Agents in Business - Deloitte '25
Competitive Landscape
Two markets exist. Neither serves mid-market. Data tools require technical teams. Automation tools are built to stitch together modern, API-native SaaS - not the legacy systems mid-market actually runs on.
Data Stack - Fragmented by Role, Not Outcome
Detail
Every tool in the modern data stack requires a technical operator. Quarri delivers business outcomes directly.
Layer
Tool
Technical Team Member Required
ELT / Ingestion
Airbyte · Stitch · Whalesync
Data Engineer
Warehousing
Snowflake · BigQuery · DuckDB
Data Engineer
Modelling
dbt · Cube · Lightdash
Analytics Engineer
Visualisation
Tableau · Looker · Omni
Data Analyst
Analytics
Hex · Python · Qlik
Data Scientist
Quarri replaces the entire stack - no technical team required.
Automation Stack - Built Around Modern SaaS, Not Legacy Systems
Detail
Workflow automation tools are built to stitch together modern, API-native SaaS - not legacy ERPs, spreadsheets, and disconnected databases. Mid-market companies run on legacy. Quarri starts from messy reality.
Quarri
Seed / Series A
Series B-C / Acquired
Series D+
Listed
Bubble size ~ last known valuation
Sources (last known round & valuation): Rows AI - Growth round ~$8.7M, May 2024 (acquired by Superhuman Feb 2026); Datarails - Series C, ~$70M, Jan 2026; $175M total; Dataiku - Series F, $200M, Dec 2022; $3.7B val; IPO expected H1 2026; Zapier - Secondary sale, ~$5B val, 2021; only $1.4M primary VC; Make - Acquired by Celonis Oct 2020 for >$100M; n8n - Series C, $180M, Oct 2025; $2.5B val; Workato - Series E, $200M, Nov 2021; $5.7B val; Tray.io - Series C, ~$50M, Sep 2022; ~$600M val; Pipedream - Series A, ~$20M, 2021; Activepieces - Seed, ~$5M, 2024; HubSpot - Listed, NYSE: HUBS; Monday.com - Listed, NASDAQ: MNDY; Airtable - Series F, $735M, Dec 2021; ~$11B val; Notion - Series C, $275M, Oct 2021; ~$10B val; Smartsheet - Listed, NYSE: SMAR; Salesforce (Agentforce) - Listed, NYSE: CRM; UiPath - Listed, NYSE: PATH; Palantir - Listed, NYSE: PLTR; Celonis - Series D, ~$1.4B total (2021-22); ~$13B val; ServiceNow - Listed, NYSE: NOW; C3 AI - Listed, NYSE: AI.
Traction · 8 Months In
$44k
ARR
$50k
Target ACV
9
Planned Pilots
4
Pipeline Partnerships
Low-margin, operationally heavy mid-market PE portfolio companies are our beachhead.
Financial controls, data accuracy, and speed directly impact portfolio P&L.
Pricing · Unit economics · GTM
Target Profile
50–500
Employees
$10–100m
Revenue
CEO / CFO
Buyer / Champion
Legacy
SaaS Systems
Target Verticals
Manufacturing & Forestry
Win Now
Hospitality & Resorts
Win Now
People Services
TBD - Future
Subscription Tiers
Essentials
$500
/month
  • 3 live sources
  • 3 users
  • Guided setup
    • Scale
    $1,200
    /month
  • 5 live sources
  • 5 users
  • 4 dev days / year
    • Business
    $2,000
    /month
  • 7 live sources
  • 10 users
  • 10 dev days / year
    • Corporate
    $3,000
    /month
  • 10 live sources
  • 25 users
  • 15 dev days / year
    • Account Expansion Model
    Chart
    Typical onboarding starts at Scale tier ($1,200/mo). Within 6 months, customers add sources, users, and professional services - driving natural expansion. Blended margin steps down as services mix increases, but absolute margin grows.
    SaaS / Professional Services Margin Contribution
    Detail
    Entry (Scale)Mid-size (Business)Large (Corporate)
    Total ACV$16.8k$36.6k$70.5k
    SaaS / Services100% / 0%75.4% / 24.6%68.1% / 31.9%
    SaaS Margin97.4%95.8%95.5%
    Services Margin-33.3%33.3%
    Blended Margin Y1 (incl. onboarding)70.7%55.8%54.4%
    Blended Margin Y2+97.4%80.4%75.7%
    Y1 includes one-off onboarding cost (data source setup + bundled dev days), absorbed in SaaS fee. SaaS margin ~95-97% - LLM and storage costs passed through. Services: contractor cost $600/day, charged at $900/day. Source: Quarri Business Plan v3.
    Unit economics model in data room
    GTM: Private Equity Distribution
    Quarri will target PE funds as a distribution channel into portfolio companies. Anthropic and OpenAI are chasing Blackstone-scale deals at the top end - leaving small-cap and mid-market PE portfolios underserved but under pressure to act.
    PE Fundraising by Fund Size (2025)
    Chart
    Small Cap
    ~1M+ boomer-owned businesses viable for sale this decade, worth up to $5T - feeding founder-led companies into PE that need professionalising for the first time
    Mid-Market
    75% of all PE deals are bolt-ons - each creating multi-entity reporting challenges across fragmented systems that need consolidating
    Large / Mega
    Anthropic, OpenAI, Palantir and major consultancies already compete here. Portfolio companies have established ERP and BI infrastructure
    Sources: McKinsey GPM 2026, PitchBook US PE Breakdown 2025, PwC Global PE Outlook 2026
    The Quarri integration has made me do more with AI in the past 6 weeks than I have in the past 6 months.
    Operations Lead
    Outdoor Recreation & Wildlife Enterprise, North America
    The Team
    Built by an operator and a data expert.
    Theo Leslie
    Theo Leslie
    CEO
    VP Growth at seed Fintech (0→$500k ARR). Dir. Strategy at Worldpay ($100m+ ARR launches). Built commercial analytics and reporting from scratch in complex, Excel-heavy mid-market operations.
    LinkedIn
    David Jayatillake
    David Jayatillake
    CTO
    3x founder (1 exit). Former VP of AI at Cube. Founded 2 semantic layer startups.
    LinkedIn
    $13.5m+ previously raised in debt + equity across both founders.
    Founder track record
    Theo Leslie - CEO
    Over a decade running processes in excel
    • VP Growth at seed Fintech/SaaS: 0 → $500k ARR
    • Dir. Strategy at Worldpay: $100m+ ARR product launches; built reporting infrastructure across a data-intensive, Excel-heavy operation
    • Founding team for delivery division at major mid-market hospitality group - built commercial analytics from scratch where manual data workflows were the structural bottleneck - and PwC
    • Over a decade frustrated by excel
    David Jayatillake - CTO
    Serial founder · data infrastructure expert
    • 3x founder (1 exit)
    • Former VP of AI at Cube
    • Founder of 2 semantic layer start-ups (one exit)
    • Leading influencer on the semantic layer
    • Data leadership roles at Lyst and Worldpay
    I can imagine tearing the current structure down and re-building it towards its actual objectives in a way that best leverages the data sources and the possible tools.
    Operations Lead
    Outdoor Recreation & Wildlife Enterprise, North America
    Vision
    Imagine a mentat in every company - on every call.
    Today
    Data foundation - connects, cleans, contextualises for humans.
    Next
    Proactive intelligence - alerts, root cause, proposes solutions.
    Future
    Agent on the call. Replaces professional services entirely.
    Moat · Stickiness
    At scale: full end-to-end automation - including professional services delivery - unlocking both SaaS and services revenue autonomously.
    The Moat
    • A more powerful data agent - targeted skill set built in data and data transformation
    • Learned tools & skills - capabilities built custom for one client can benefit all clients
    • Tight workflow tailoring to mid-market - tools, context and solution layer built around legacy mid-market challenges
    Embedded Stickiness
    • Growing memory - every interaction makes the system smarter about your business
    • Infrastructure lock-in - embedded in data infrastructure, hard to rip out
    • Operational embedding - data capture, reconciliation, and operational workflows become business-critical. The deeper Quarri gets into daily operations, the harder it is to replace
    The Ask
    $2.4M
    18 months runway.
    First raise. Founder-funded to ARR.
    Forecast · Use of funds · Roadmap
    Raise Timeline
    First Meetings
    April – May
    Lead Term Sheet
    May – June
    Close
    Q2/Q3 2026
    Financial Projections
    CY2026CY2027CY2028CY2029
    Total Revenue$61k$2.4M$11.7M$35.2M
    Exit ARR$231k$5.1M$20.2M$49.5M
    Gross Margin76%61%63%64%
    Total Contracts51655421,096
    Headcount*6162739
    Rev / Employee$10k$152k$431k$904k
    Source: Quarri Business Plan v3. SaaS/Services split: CY2026 78%/22%, CY2027 81%/19%, CY2028 78%/22%, CY2029 74%/26%.
    GM compresses CY2027 as new-customer onboarding peaks; recovers as cohorts mature. Per-customer steady-state GM 75–95% — see Slide 8 unit economics.
    * Headcount reflects roles, not literal heads - we expect to fill some through AI agents rather than hires as the platform matures.
    Financial forecast in data room
    Product (18 months)
    • Optimise for Claude as front end
    • Deepen analytical functionality & context memory
    • Build bank of reusable workflows across our clients
    • Use meeting notes to create an AI-powered implementation agent to replace ‘custom dev days’
    Hires*
    • BD1 — Aug 2026 (M5)
    • Midlevel Developer — Aug 2026 (M5) · GTM & auto-onboarding focus
    • SDR1 / GTM hire — Sep 2026 (M6)
    • Senior Developer — Nov 2026 (M8)
    • Junior Developer — Jan 2027 (M10)
    Two founders (CEO, CTO) full-time from Day 1. CY2026 EOY headcount = 6 (2 founders + first 4 hires; Junior Developer joins early CY2027). Marketing budget deployed from month 1.
    * These are roles, not literal heads - some may be filled by AI agents rather than hires.
    Current Ownership
    David Jayatillake 50% · Theo Leslie 50%.
    ARR Milestones
    Chart
    Use of Funds Breakdown
    Chart
    Get in Touch
    Theo Leslie
    Theo Leslie
    Co-Founder & CEO
    theo@quarri.ai · quarri.ai
    LinkedIn