LUMORAProject Lantern — Executive diligence dashboard · Prepared by Yajur Partners

Overview

Illustrative sample · anonymisedASIS · Management CaseUSD MnHorizon FY22A–FY31P

Investment overview

A founder-funded, EBITDA-positive Education OS compounding across two engines

Lumora operates a curriculum-native school transformation and language-learning platform: an India B2B private-school engine at inflection and a Canada B2G cash engine with 11-year ministry tenure. Every figure ties to the management financial model dated 03 Sep 2026 (consolidated, FY ending 31 March); operational and market datasets are labelled. Click a tile to jump to its detail.

Operating revenue by geography

Operating revenue (excl. other income). FY22A–FY26A actuals; FY27E; FY28P–FY31P. International = Canada + Africa from FY28P. New revenue lines are embedded within geographies.

Adjusted EBITDA

Adjusted for promoter items, prior-period discount adjustments and other income (FY22A–FY26A only; no adjustments from FY27E). Margin on operating revenue shown in tooltip and in the summary table below.

Financial health heatmap — the diligence grid at a glance

low → high within each row (blue = favourable direction) unfavourable / negative values

Equity story — what a control investor underwrites

Diligence friction, pre-answered

Diligence scorecard YP assessment

Harvey-ball ratings are Yajur Partners' assessment against buyout-fund underwriting norms for institutional edtech; each is anchored to the evidence cited and to the tabs of this dashboard. Ratings do not change with the scenario toggle.

Scenario divergence — indexed to FY26A = 100

Operating revenue and adjusted EBITDA for both cases on one index scale. ASIS (Management Case) is the underwriting base; Growth is upside (faster India school adds, larger Canada step-up, deeper Africa rollout, faster new-line attach).

Severability check — what FY30P looks like without the unproven pieces Illustrative overlay on ASIS

Haircut to India core growth FY26A → FY30P: 0%
Haircut to Canada growth FY26A → FY30P: 0%

Method: removes the selected FY30P revenue from the ASIS case and flows the loss through at the model's own FY26A→FY30P incremental adj. EBITDA margin (). Not a re-run of the model — an arithmetic severability test on the management case, the way a deal team stress-tests a CIM. Africa-market and new-line revenue may partially overlap in the model's geography split, so excluding both together is deliberately conservative.

Headline financial summary

Financial metrics

Growth, margins and the ratios a fund computes first

Scenario, currency and horizon follow the header controls. Segment view toggles geography for the charts. Balance-sheet-derived ratios are computed from the model's full three-statement build, available for both scenarios in the 03 Sep 2026 model.

Adj. EBITDA bridge FY26A → FY30P — volume vs margin

Volume effect = Δrevenue × FY26A adj. margin; margin-expansion effect = FY30P revenue × Δmargin. Separates "more of the same" from operating leverage.

Revenue bridge FY26A → FY30P

Operating revenue by geography. New revenue lines (VoxLearn, HPC Labs, SchoolMIS, RoboLabs, AI Labs, ExamPrep, FLN Suite) sit inside the geography increments.

Segment heatmap — growth and margin by geography and year

low → high within each row negative
Segment view

Operating revenue

EBITDA & margin

Segment EBITDA per the model's geography allocation (reported basis; FY22A–FY26A adjustments are consolidated-only).

Margin structure — consolidated

Revenue composition — existing vs new lines

Solutions & resources = platform + content subscriptions (recurring). Hardware & spares run-off is 0.7% of FY26A revenue. New lines reach $26.4M by FY31P even in ASIS ($79.8M Growth).

Degree of operating leverage

DOL = %Δ(adj. EBITDA − D&A) ÷ %Δrevenue, from FY26A where the base is positive and stable. Flow-through (Δadj. EBITDA ÷ Δrevenue) in tooltip.

Rule of 40 — revenue growth + adj. EBITDA margin

Operating-revenue growth stacked on adj. EBITDA margin; marker at 40. Clears 40 from FY27E in the ASIS case and in every projected year of the Growth case (history is identical in both).

Interpretation panel — ratios by year

Derived from the statement tables. Cash-conversion cycle = DSO + DIO − DPO (direct-cost basis). Incremental margin = Δadj. EBITDA ÷ Δrevenue. Interest cover = adj. EBITDA ÷ interest paid. Pre-tax ROCE = (adj. EBITDA − D&A) ÷ (equity + net debt); not meaningful once capital employed turns to net cash (FY29P+).

New revenue lines — product build

Seven product lines on the same platform core, FY27E–FY31P. India leans VoxLearn / RoboLabs / AI Labs; International leans HPC Labs / SchoolMIS. Gross margins per line are model-derived (revenue less direct cost).

Segment metrics — full series

Financial statements

Three-statement model — consolidated, FY22A–FY31P

Full P&L, balance sheet and cash flow from the 03 Sep 2026 management model, both scenarios. FY23A–FY25A audited (FY22A per model); FY26A management-certified, provisional (audit commences Q2 FY27); FY27E–FY31P projections. The consolidated statements in the workbook (FY23–FY25 audited, FY26 provisional) are reproduced, with a line-by-line tie-out to the model. Download the workbook from the header to work with the numbers directly.

Profit & Loss

Below-EBITDA profile

Finance cost declines from $1.15M (FY26A) to $0.72M (FY31P) as term debt amortises (Lender A LAP/demand loan, Lender B, SBA, vehicle loans). D&A steps down as capitalised content completes FY29P.

Model integrity checks

Checks recomputed in-page from the embedded model data on every toggle. Tolerance ±$0.01M.

Operations & unit economics

Learners, schools, cohorts, footprint — and what each one earns

Model-basis learner and school counts (both scenarios) alongside management operational data for India (release cohorts, states, zones, concentration; FY22-23 → FY25-26) and the model's CAC / CLTV build.

Learners by segment and grade band

Schools by geography

India FY26A 1,281 schools (model basis, identical to the IM distinct-name count). Operational cohort tables carry the school-level file's 1,342 unique-account basis; learners and INR values are identical on both bases.

India unit metrics — implied from the model, and fee inputs

Implied ARPU = India operating revenue ÷ India learners; revenue per school = India operating revenue ÷ India schools. Fee inputs are the model's price assumptions by grade band (India in ₹ / learner / year; International in $ / learner / year). FY26A: ~$20 ARPU, ~$6.5K per school on ~322 learners per school. International: fee inputs × learners do not reconcile to International revenue ($2.4M vs $4.3M FY26A; implied ARPU $285) — the Canadian contracts include programme fees beyond per-learner pricing; shown in the segment table and a management confirmation item.

Release cohorts — India schools by fiscal year Operational data

Year-on-year retention — schools and contract value Operational data

Blended count retention was 80% (FY23-24), 78% (FY24-25) and 49% (FY25-26, the R5 launch year); value retention 57%, 64% and 53%. The model build carries 49–60% for FY24–FY26 and holds India at 50% from FY27E; management states ~60% renewals. Release rows in the cohort table sum to slightly more than the unique-school totals (e.g. 2,066 vs 2,049 cumulative) where schools moved between releases — to confirm with management.

Release cohorts — learners & contract value Operational data

Zone mix FY25-26 — contract value

India customer concentration — top 10 accounts, FY25-26

India footprint — state heatmap, FY25-26

State-wise footprint — schools, learners, value (₹), FY22-23 → FY25-26

Click headers to sort. Grand totals FY25-26: 1,342 schools · 411,788 learners · ₹85.4 Cr. Top-5 states by value (TG, AP, KA, MH, UP) hold ~56% of contract value; Telangana alone 20%. No single school >7.6% of India revenue.

Unit economics — CAC, CLTV and payback

India — CAC vs CLTV per learner (USD) CAC/CLTV build

India — LTV/CAC ratio

FY25's 5.0x reflects a CLTV spike (pricing mix); the build shows 4.4x FY26 rising to 6.3x FY29P and 7.9x FY31P as CAC falls. Canada LTV/CAC runs ~68x (FY26) given ministry-level acquisition economics. Source: management model CAC/CLTV build (fully allocated sales, inside sales, customer success, conveyance and S&M; not part of the abridged model).

Unit stack — India, per learner, FY26 (USD)

ARPU from the model (India revenue ÷ learners); gross profit per learner at the CAC/CLTV build's India gross margin (~57%); CAC and CLTV from the build. Gross profit per learner (~$13.7) recovers the $10.9 CAC in about 9.5 months.

Retention assumptions by geography

India deal value per school

Manpower plan FY30 — ASIS (404 heads) Mgmt plan

Current headcount 370. ASIS reaches $48.6M revenue with ~404 heads (~$120K per head); the Growth-case org plan scales to 1,212 heads for the larger book. Hiring gated to measurable scale outputs.

Balance sheet, cash flow & debt

Cash generation, working capital and the debt profile

Consolidated, from the model's balance sheet and cash-flow build (both scenarios). FY22A–FY25A audited; FY26A management-certified; FY27E+ model. Working-capital and funding characteristics are presented with their operating context.

Cash conversion funnel

Adj. EBITDA → cash from operations → free cash flow → FCFE after interest and scheduled debt service, for the selected year. The model's own FCFE line is before interest (interest sits in financing); the bridge table shows both.

EBITDA-to-FCF conversion path (%)

ASIS: FCF ÷ adj. EBITDA inflects FY27E (31%), reaches 74% FY28P and ~90% from FY29P (Growth: 25% / 53% / 69%, then 97%+). Drivers: receivables normalisation, content capex completing FY29P, and other current liabilities rising to 52% of revenue by FY31P (63% Growth) — see the funding composition chart. Unlevered structural measure (adj. EBITDA − capex) ÷ adj. EBITDA exceeds 90% from FY29P.

Gross and net debt

Gross debt $13.0M at FY26A (incl. $1.0M interest-free promoter / related-party loans) vs $0.17M cash → 3.5x net debt / adj. EBITDA (4.1x on reported EBITDA). Amortisation only — no refinancing or equity assumed. Model cash is negative in FY27E (−$0.3M ASIS, −$0.6M Growth) while $0.8M of debt is repaid; the Lender A OD is fully drawn, so working-capital headroom is a management confirmation item. USD balances are translated at the model's average FX per year.

Net debt / adj. EBITDA

Deleverages to ~1.0x by FY28P (ASIS; 0.7x Growth); net cash from FY29P in both scenarios. On reported rather than adjusted EBITDA, FY26A leverage is 4.1x.

Cash from operations and free cash flow

FCF = CFO − capex (fixed assets + capitalised content/intangibles). Content-build capex ends FY29P. FY23A–FY26A FCF suppressed by the deliberate content build and B2G receivable growth.

Working capital — receivables and DSO

DSO peaks at 316 days in FY26A. Receivables ($12.3M) exceed total International revenue ($4.3M), so at least $8.0M is owed by India private schools — the ageing by geography is the first data-room request. The model normalises to 223 days FY27E and 128 by FY31P. Receivable balances in the detail table.

Balance sheet composition — assets

Balance sheet composition — funding

Other liabilities (deferred revenue / advances and provisions) grow from $2.3M (16% of revenue, FY26A) to $38.6M (52%, FY31P) in ASIS and supply ~$36M of the ~$78M cumulative FY27E–FY31P free cash flow — the largest single working-capital assumption in the model. Equity compounds from retained earnings alone; no equity infusion is modelled.

Adj. EBITDA → free cash flow bridge & conversion ratios

FCF = CFO − capex; conversion = FCF ÷ adj. EBITDA. The working-capital line is defined as CFO − adj. EBITDA − taxes paid so the bridge foots exactly (it therefore also carries other-income timing and non-cash items). Pre-tax ROCE shown through FY28P while capital employed is positive.

Debt schedule — loan by loan As at 31 Mar 2026

Market, competition & exit

The runway, the field and the exit environment

Verified market figures only (UDISE+ / Economic Survey basis); estimates labelled. Competition framing per buyer Q&A responses; benchmark financials from filings via Tracxn (4 Aug 2026). Exit datapoints per public reporting to Aug 2026.

India penetration path — learners vs the K-5 private segment

Even at FY31P (~1.04M learners, ASIS) Lumora reaches ~2.3% of the ~46M K-5 private segment; Growth case ~2.3M (~5.0%). The plan consumes a sliver of the runway.

The Scaled Peer A contrast

Scaled Peer A figures per public filings; capital raised ~$185M per public reporting. Lumora reached profitability on zero external equity.

Exit environment — the public market has reopened for profitable edtech

Why now — four converging forces

The second engine — Africa & the Canada mandate

Sources: national education ministry statistics, World Bank/UNESCO UIS 2024, provincial assessment consortium statements & Jan-2026 results release. Regional totals are sums across mixed vintages — estimate.

Canada revenue concentration, FY25-26

Structural to B2G: three provincial ministries (Province A 32%, Province B 20%, Province C 16%) + the platform partner stream (25%) each exceed 15% of regional revenue, 93% combined. Contracts are multi-year with auto-renewal provisions; India is the diversified counterweight (largest account 7.6%).

Competitive benchmark — disclosed financials

Competitor figures verified from filings via Tracxn; reach figures company-reported. 36.5% of Indian schools have no working connectivity (UDISE+ 2024-25) — Lumora's offline-first architecture is the only full stack that operates there.

Competitive landscape — India K-12

Global & regional landscape

Q&A bank & basis of preparation

Every question funds have asked — with the reconciled answer

Compiled from investor FAQs and buyer diligence queries received to date, restated on the 03 Sep 2026 model. Filter by theme or use the header search. Click a question to expand.

Basis of preparation

Ask the dashboard

Ask a question — get an answer with the chart behind it

Type any question about Lumora's financials, operations, unit economics, market or the transaction. The assistant reads the same model data, Q&A bank and basis notes that power every tab, answers in prose, and draws the chart(s) that support the answer. Quantitative questions are computed live from the embedded model; qualitative ones draw on the diligence Q&A.

Try asking

How it answers Built-in engine

The built-in engine recognises metrics (revenue, EBITDA, margins, PAT, cash, debt, DSO, learners, schools, ARPU, CAC…), segments (India, Canada, Africa), years and scenarios, computes the answer from the model and renders the chart. For "why / how / what is" questions it retrieves the closest answers from the 46-question diligence Q&A.

When the site is deployed with the optional Claude connector (Netlify function + API key, see README in the deploy folder), questions are answered by Claude with the full dashboard context and richer reasoning; the built-in engine remains the fallback.

Project Lantern · Illustrative sample · Anonymised demonstration dashboard — company name, product names, counterparties and all financial figures have been changed and do not represent any actual business · Prepared by Yajur Partners