Rogo
B2B enterprise SaaS - per-seat subscriptions (~$3,300/seat/yr per Sacra estimate) plus enterprise deployments; land-and-expand across banks, PE firms, hedge funds, and asset managers
Series B, C, and D marks are press-reported post-money valuations from priced primary rounds. The Series A point is an estimate band (valuation undisclosed; shown at $0.08B for chart continuity - treat as indicative only). No known secondary or SPAC transactions.
Earnings, margins, COGS & capex
Rogo does not publish audited financials. Bloomberg (Oct 2025) reported revenue of ~$2M in 2024 rising to $15M+ in 2025; the OpenAI case study (early 2025) cites 27x ARR growth, and the company reports its Felix agent reached seven-figure ARR within five months of launch with a single sales rep. The Series D valuation (~$2B post) against press-reported eight-figure revenue implies a very high revenue multiple typical of top-decile 2025-26 vertical-AI rounds. Total capital raised exceeds $300M: seed backing from AlleyCorp, BoxGroup, Company Ventures, and ScOp (pre-2024, company emerged from stealth early 2024); Series A $18M Oct 2024 led by Khosla Ventures (Fortune reported $18.5M); Series B $50M Apr 2025 led by Thrive Capital at $350M post; Series C $75M announced Jan 2026 led by Sequoia at $750M post; Series D $160M Apr 2026 led by Kleiner Perkins at ~$2B post.
Revenue trend
Margins
LLM inference costs are the key swing factor; multi-model routing (fine-tuned GPT-5, Gemini 3, smaller reasoning models per Sacra) suggests active cost optimization
heavy investment in engineering headcount and post-Series C European expansion
COGS structure
Dominated by LLM API inference (OpenAI, Google, Anthropic) and cloud compute, plus licensed financial data feeds (S&P Global, Crunchbase, FactSet cited in the OpenAI case study) and human review - former bankers label and QA datasets - a services-heavier COGS profile than pure SaaS.
Capex
Minimal traditional capex; compute consumed as opex through model providers rather than owned infrastructure. Capital deployment is concentrated in R&D headcount and go-to-market.
Latest earnings
n/a
none published
- Last priced round
- $160M Series D at ~$2B post (Kleiner Perkins, announced 2026-04-29)
- Total raised
- $300M+
- Users / institutions
- 35,000+ professionals / 250+ institutions (company-reported, Apr 2026)
- Estimated per-seat pricing
- ~$3,300/yr (Sacra estimate)
- Headcount
- not disclosed (NY HQ; European expansion announced Jan 2026)
Growth drivers
- Land-and-expand inside bulge-bracket and elite-boutique banks — 35,000+ professionals at 250+ institutions (Rothschild & Co, Jefferies, Lazard, Moelis, Nomura, Raymond James) as of the Apr 2026 Series D, up from 25,000 users at 150 firms in Oct 2025
- Expansion from investment banking into private equity and buy-side (GTCR, Siris Capital, Tiger Global as customers)
- Geographic expansion: European push announced with the Jan 2026 Series C
- New product surface — Felix agentic AI (deal screening, CIM generation, buyer outreach, data-room diligence) and distribution as a launch partner on Anthropic's Claude Marketplace (Mar 2026)
- Deep firm-data-warehouse and Office-suite integration raising switching costs
Bull & bear
Rogo is the emerging system-of-work for high finance: it owns the junior-banker workflow inside the world's most reference-driven buyer network, is growing revenue at multiples per year, and has assembled a capital-and-distribution syndicate (Sequoia, Kleiner Perkins, Thrive, Khosla, JPM, Wells Fargo, Kravis) that competitors cannot match. If it converts 35,000 seats into deep multi-product enterprise contracts, today's ~$2B mark looks cheap.
- Category leadership signal: three rounds in 12 months, each roughly 2-3x the prior valuation, with top-tier leads (Thrive -> Sequoia -> Kleiner Perkins) doing diligence on real usage data
- Workflow lock-in: embedding in Excel/PowerPoint/Word plus firm data warehouses creates switching costs terminals never had
- Seat economics can inflect: 35,000 users at even $5-10K blended enterprise pricing implies a $175-350M ARR ceiling in the existing installed base alone before new logos (scenario math, not disclosed)
- Buyer network effects: banking is a prestige-reference market; Rothschild/Lazard/Jefferies logos compress sales cycles for the next 250 institutions
- Felix agentic expansion multiplies revenue per account: seven-figure ARR within 5 months of launch with one sales rep (company-reported), plus expansion into PE, credit, and asset management beyond the IB core
Rogo is a thin application layer on top of models it does not own, selling into a small, concentrated buyer universe, at a valuation (~$2B) that is 100x+ its Bloomberg-reported 2025 revenue. Model providers and data incumbents are converging on the same workflow from above and below, and the integration-heavy deployment model may never produce software margins.
- Valuation risk: ~$2B post against press-reported $15M+ revenue (2025) leaves no room for growth deceleration; a flat or down round would damage the prestige-sensitive customer narrative
- Platform risk: OpenAI/Google/Anthropic capture the reasoning value and can raise inference prices or ship competing finance agents; Rogo's own OpenAI and Google Cloud case studies advertise the dependency
- Moat question: prompts, connectors, and Office integrations are replicable; incumbents (Bloomberg, FactSet, S&P) own the proprietary data Rogo must license or access via clients
- Buyer concentration and procurement power: a few hundred institutions with strong vendor-squeeze leverage; bank in-house AI builds are an explicit substitute
- Human review teams (former bankers labeling data) and deep per-client integration scale with revenue - a Palantir-style margin profile without Palantir-scale contracts yet
What it is worth
Last priced round + venture revenue-multiple sanity check (no DCF possible - financials undisclosed)
$700M-1B repricing
if growth decelerates below 2x, model providers ship competing finance agents, or the AI multiple regime resets - a down round below the Series C mark would also impair the customer-facing narrative
~$2B mark holds
company grows into 20-30x a $70-100M ARR by 2027 and raises flat-to-up
$4-6B next-round scenario
if ARR reaches $100M+ with bank-grade net retention and European/buy-side expansion lands (Harvey-style trajectory)
~$2B post-money (Series D, Kleiner Perkins, announced 2026-04-29) is the only market-cleared mark. Against Bloomberg-reported $15M+ revenue (2025) that is a 100x+ trailing multiple; against a plausible 2026 run-rate in the $40-80M range (extrapolated from growth citations and 35,000 seats - NOT disclosed, treat as scenario math), it is 25-50x forward - top of the 2025-26 vertical-AI band (comparables: Harvey, Hebbia, Glean rounds). The mark prices continued 3x+ annual growth and successful buy-side expansion. Not financial advice; no public shares exist and secondaries are illiquid.
SWOT
Strengths
- Deep vertical focus on high finance with purpose-built financial reasoning models and workflows (Excel/PowerPoint/Word integration), not a generic chatbot
- Marquee reference customers (Rothschild & Co, Jefferies, Lazard, Moelis, Nomura) in a reference-driven buyer market
- Investor syndicate doubles as distribution and validation — J.P. Morgan Growth Equity Partners, Wells Fargo, Henry Kravis (KKR co-founder), and Tiger Global - several investors are also customers
- Rapid valuation momentum ($350M -> $750M -> ~$2B in 12 months) gives capital advantage over smaller rivals
- Founder-market fit — CEO Gabriel Stengel is an ex-Lazard banker (Princeton CS) building for his former workflow, co-founded with John Willett and Tumas Rackaitis
Weaknesses
- Revenue base (Bloomberg-reported $15M+ in 2025) is small relative to a ~$2B valuation - execution must stay near-perfect to grow into the multiple
- Dependent on third-party frontier models (OpenAI, Google, Anthropic) it does not control for core product quality and cost
- Human-in-the-loop data labeling and deep per-client integration are services-heavy and scale with headcount, pressuring gross margin
- Long, compliance-heavy enterprise sales cycles at banks — concentration risk among a modest number of large institutions
Opportunities
- Bank IT budgets shifting from terminals/data licenses (Bloomberg, FactSet, Capital IQ) toward AI-native workflow tools - a multi-billion-dollar wallet to attack
- Agentic automation of junior-banker work (deal screening, CIM generation, deck assembly) is still early; whoever owns the workflow owns the seat
- European expansion off a US beachhead; PE/credit/asset-management segments less penetrated than banking
- Marketplace distribution (Claude Marketplace launch partner alongside Snowflake, GitLab, Harvey) and potential platform partnerships with model providers
Threats
- Foundation-model providers moving up the stack — OpenAI, Anthropic, and Microsoft Copilot shipping finance-tuned agents could commoditize the application layer
- Incumbents with data moats (Bloomberg, FactSet Mercury, S&P Capital IQ, AlphaSense, Moody's) bundling AI into existing contracts
- Well-funded direct rivals (Hebbia, BlueFlame AI, Brightwave) and banks' in-house LLM builds (e.g., internal GPT deployments at JPMorgan, Goldman)
- AI-valuation regime reset would make the next round punitive given the ~$2B mark against press-reported eight-figure revenue
Moats, dependencies & bottlenecks
Moats
Excel/PPT/Word integration and firm-data-warehouse connectors make rip-out painful once adopted firm-wide
Rothschild & Co, Lazard, Jefferies, Moelis, Nomura logos are the strongest sales asset in banking software
Fine-tuned frontier models plus banker-labeled datasets (per OpenAI case study); durable only if usage data compounds faster than frontier models generalize
Passing InfoSec/model-risk review at 250+ regulated institutions is a real barrier to entry for startups, less so for incumbents
Dependencies
technology (frontier models) Published case study cites o1/o1-mini/GPT-4o powering research and 27x ARR growth (early 2025); Sacra cites fine-tuned GPT-5 for high-stakes analysis; provider could compete directly or reprice
Gemini 3 in the multi-model stack per Sacra; Google Cloud publishes its own Rogo customer case study
technology + distribution Claude Marketplace launch partner (Mar 2026, alongside Snowflake, GitLab, Harvey, Replit, Lovable)
Product lives inside Excel/PowerPoint/Word; Microsoft Copilot is simultaneously the platform owner and a competitor (MSFT)
FactSet, Crunchbase, client-licensed feeds) Analyst-grade output requires licensed market/filings data Rogo does not own (S&P Global, Crunchbase, FactSet named in OpenAI case study); licensors are also competitors (SPGI, FDS)
Standard hyperscaler dependency for hosting and inference
Advantages
- First-scaled mover in AI-native investment-banking workflow with the deepest logo roster in the category
- Capital superiority — $300M+ raised, ~$2B mark, and strategic investors who are also customers/channels (J.P. Morgan, Wells Fargo, Tiger Global, Kravis)
- Multi-model architecture (fine-tuned GPT-5, Gemini 3, smaller reasoning models per Sacra; Claude via marketplace) reduces single-vendor exposure and enables cost/quality routing
- Founder domain credibility (ex-Lazard CEO) in a buyer market that distrusts generic tech vendors
Weaknesses
- Application-layer dependency on rivals' foundation models
- Revenue small relative to valuation; multiple assumes years of hypergrowth
- Services-heavy gross-margin profile from human review and per-client integration
- Narrow buyer universe (~few hundred institutions) with strong procurement leverage
- No public financial disclosure - outside investors cannot verify unit economics
Bottlenecks
- Enterprise sales + model-risk/compliance review cycles at regulated banks gate deployment speed
- Banker-grade talent for data labeling, QA, and client integration - the deployment model is headcount-constrained
- LLM inference cost and latency at scale across 35,000+ seats
- Access to licensed proprietary data (Bloomberg/FactSet/Capital IQ content) that clients hold but Rogo cannot redistribute
Top signals & trends
Top signals
Preemptive round pattern typical of breakout vertical-AI leaders; also raises down-round risk if growth slows
Strategic-buyer money doubles as validation and channel
Seat count up sharply; monetization depth per seat is the open question
New enterprise distribution, but deepens platform dependency on a model provider
Marketing wins that simultaneously document the platform-risk bear case
Early evidence agentic products expand revenue per account; company-reported and unaudited
Trends
strong tailwind · Deal screening, CIM generation, buyer outreach, and data-room diligence are exactly Felix's stated automation targets
OpenAI/Anthropic/Microsoft finance agents could compress the application layer Rogo occupies
Puts Bloomberg/FactSet/Capital IQ budgets in play for AI-native challengers
Raises deployment friction but rewards vendors who have already passed 250+ institutional reviews
tailwind now, risk later · Cheap capital today; brutal comparables if the regime resets before revenue catches up
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Frontier reasoning models (o-series, GPT-4o, fine-tuned GPT-5 per Sacra) powering core research workflows
Gemini 3 in the multi-model stack; Google Cloud infrastructure (published Rogo customer case study)
Claude models + Claude Marketplace distribution
Office platform (Excel/PPT/Word integration) and Azure-hosted OpenAI capacity
Licensed financial data (named in OpenAI case study) underpinning analyst-grade outputs
Licensed data feeds (named in OpenAI case study) Rogo integrates against
Listed on Euronext Paris (ROTH.PA); flagship advisory-bank deployment
Full-service investment bank customer; published a leadership spotlight on Rogo's CEO
Elite advisory firm; CEO Stengel's former employer
Independent advisory bank customer
Japanese global bank (NYSE ADR); anchor for international expansion
Private equity firm customer - evidence of buy-side expansion (alongside Siris Capital and Tiger Global)
Closest private rival; AI document-analysis platform (Matrix) for asset managers/banks; $130M Series B led by a16z at ~$700M valuation (Jul 2024)
Private market-intelligence/search platform (~$4B valuation, 2024 round) adding generative features; owns content licensing Rogo lacks
Copilot embedded in the same Office surface Rogo lives in; platform owner and competitor
Model supplier and potential direct competitor via finance-tuned ChatGPT/agents for enterprises
Private; Terminal incumbency plus BloombergGPT and AI document search across the buy side and sell side
Mercury conversational AI and deep workstation entrenchment in banking workflows
Capital IQ Pro with generative AI features plus Kensho; owns the underlying data
Research Assistant (GenAI) across credit and research content
Private; LLM-agnostic AI platform focused on alternative-investment managers
Private; AI research agent for investment professionals
Model supplier and marketplace partner shipping Claude for Financial Services - partner today, potential competitor tomorrow