Disha Project Audit Report

Consolidated Findings Across All 33 Documents


CRITICAL DATA FIXES (Must fix before any external presentation)

Data Inconsistencies

# Issue File A Value File B Value Severity
1 PLI jobs 04-govt-policy 28,884 05-china-plus-one 14.39 lakh CRITICAL — 50x gap
2 Internet access 02-supply-side 15% households 06-ai-disruption 806M users, 85.5% smartphones CRITICAL — outdated data
3 CSR education spend FY23 04-govt-policy Rs 10,085 Cr (33%) 07-labor-mobility Rs 13,209 Cr (44%) HIGH
4 NAPS stipend structure 04-govt-policy 25% / Rs 1,500 cap 09-gtm-research 50% / Rs 4,500 cap HIGH — pre vs post reform
5 GER current 01-demand / 04-policy 28.4% 02-supply-side 32.5% HIGH — different years
6 NSQF levels 02-supply-side 8 levels 07-labor-mobility 10 levels MEDIUM — 07 is outdated
7 University count 01-demand 1,168 02-supply 1,338 MEDIUM — data year
8 ITI utilization 02-supply 48% 05-china-plus-one 42% MEDIUM
9 Allen revenue FY24 08-ops-research Rs 3,244 Cr 09-gtm-research Rs 3,473 Cr LOW
10 GCC count 00-framework 1,900 03-employer 1,800+ LOW

Suspicious/Likely Wrong Numbers

Claim File Problem
"35M STEM graduates/year" 05-china-plus-one Almost certainly wrong — total higher ed enrollment is 43M. Annual graduates would be ~10M total, STEM fraction even less
"10M direct + 50M indirect EV jobs by 2030" 03-employer 60M total from 150K current = 333x in 5 years. Needs source verification
"115M total new jobs by 2030" 05-china-plus-one 7.85M/year x 5 = 39.25M, not 115M. Math doesn't add up
"7M new AHP jobs annually" 03-employer Total shortage is 6.5M. Annual need exceeding total gap makes no sense unless includes replacement demand — unstated
Education loan NPA "~4%" 00-framework Source says 2% gross NPA (PSBs) and ~10% historical default. 4% appears interpolated

Structural Issues

  • README file names are ALL wrong — every filename in the README table doesn't match the actual files
  • Files 08 and 09 missing from framework index (00-structural-forces-framework.md)
  • No source citations in any of the 10 research files — hundreds of claims with zero formal references

FRAMEWORK GAPS

Missing Frameworks (4 identified)

# Framework Why it matters
1 Competitive Landscape No mapping of competitors (NIIT, Aptech, TeamLease, Masai, Generation India). No competitive response scenarios
2 Team & Talent Strategy Founding team requirements, first 10 hires, trainer recruitment pipeline, compensation philosophy — all absent
3 Financial Model & Funding Strategy No per-center P&L, no sensitivity analysis, no fundraising timeline, no investor targeting
4 India Stack Integration Aadhaar KYC, DigiLocker credentialing, ONEST discovery, ABC credit banking, UPI payments — all mentioned but never architected

Internal Coherence Issues

Issue Details
Vernacular timeline conflict Moat doc (06) says 12-18 months to defensibility per language. Scaling doc (07) expects 3 languages in 6 months.
Revenue inconsistency at scale Phase 3: 20K students × Rs 50K = Rs 100 Cr (below stated Rs 150 Cr floor). Phase ranges don't align at extremes
Flywheel missing Proximity pillar Trust equation (01) has 4 pillars including Proximity. Flywheel (04) operationalizes 3 of 4 but never addresses Proximity
Year 2/5 revenue mismatch Doc 07 says Rs 14 Cr (Y2), Rs 180 Cr (Y5). Doc 08 says Rs 13.2 Cr (Y2), Rs 195 Cr (Y5). Neither specifies gross vs net

Critical Strategic Gaps

  1. Cold start problem — How to fill Cohort 1 with zero brand, zero placement data. Mentioned but never strategized
  2. Employer demand validation — The entire model rests on employers co-designing, pre-committing, and paying placement fees. Never stress-tested against evidence
  3. Gender strategy — Acknowledged as important but operationally thin. No specific interventions for enrollment, safety, childcare, marriage penalty
  4. AI disruption scenario — What if AI eliminates entry-level cloud jobs (the primary revenue track) within 3-5 years? One-line mitigation only

OPERATIONS GAPS

Missing Core Documents (4 critical)

# Document Why it's essential
1 Daily Operations Manual No runbook for how a center actually runs day-to-day
2 Student Lifecycle Document No end-to-end journey: inquiry → enrollment → training → placement → alumni
3 Legal/Contracts Framework No templates for: student agreements, ISA contracts, employer MoUs, trainer contracts, co-location leases
4 Financial Operations Plan No cash flow management, working capital requirements, collections process, AR management

Number Mismatches Across Ops Docs

Issue Doc A vs Doc B
Hub staffing Doc 01: 15-25 FTEs vs Doc 02: 14-17 FTEs
Spoke staffing Doc 01: 4-8 FTEs vs Doc 02: exactly 5
Hub monthly opex Doc 01: Rs 8-12L vs Doc 02: Rs 6.5-10L (staff only) + rent = Rs 8.5-13L
Trainer salary Doc 04: Rs 50-80K/mo vs Doc 07 implies Rs 60-90K/mo
PMKVY subsidy Doc 07/08: Rs 10-20K vs Research 08: Rs 28,900-40,400 (actual PMKVY 4.0 rates)
Year 5 revenue Doc 07: Rs 180 Cr vs Doc 08: Rs 195 Cr

Missing Operational Topics

  • Safety & insurance — Manufacturing (CNC, EV) and healthcare (clinical) tracks have zero safety protocols, no student insurance, no liability framework
  • Data privacy — DPDP Act mentioned once, no implementation plan, no DPO role, no consent management
  • Crisis/BCP — No pandemic plan, no trainer mass-attrition contingency (FIITJEE scenario), no reputational crisis playbook
  • Entity structure — Section 8 vs Private Limited unresolved, cascades through tax, CSR eligibility, govt access
  • FIITJEE lesson not echoed — The #1 cautionary tale (collapsed from unpaid salaries) doesn't appear in org design, trainer model, or unit economics docs where it matters most

Track-Specific Gap

Doc 04 uses uniform 1:30 trainer ratio across all tracks. Research shows manufacturing and healthcare require 1:10-15 for safety/clinical reasons. This is dangerous.


EXPERT QUESTIONS GAPS

Missing Expert Types (Critical)

# Expert Type Why needed
1 Students & parents The actual customer. All psychographics are assumptions, not validated. No interview guide exists
2 NBFC/lending partners ISA structured through NBFCs is a key revenue assumption. No validation path
3 ITI/polytechnic principals Co-location strategy depends on them. Different from "policy people in Delhi"
4 Healthcare sector experts Clinical practicum access, hospital partnerships, allied health regulations
5 International mobility experts POE registration, bilateral program realities, emigration regulations

Load-Bearing Assumptions with ZERO Interview Coverage

Assumption Source Validated by?
AI tutor handles 80% of queries Trainer Model NONE
Clinical practicum slots at hospitals Curriculum NONE
Co-location with ITIs viable Center Architecture NONE
Vernacular content production cost Moat Architecture Partially (1 question)
ISA collection rate 80-85% Revenue Model Partially (1 question)
30%+ referral enrollment by Month 18 Flywheel Tangentially

Question Quality Issues

  • 5 questions too broad (will get vague answers) — specifics noted in full audit
  • 2 questions are leading (give thesis, ask expert to agree)
  • Several missing specificity anchors (no city tier, student profile, center size specified)
  • No interview sequencing guide exists

Recommended Interview Sequence

  1. Education Operators (Weeks 1-3) — ground truth on costs and operations
  2. Edtech Founders (Weeks 1-3) — validates unit economics and completion rates
  3. Employers (Weeks 3-5) — demand validation, the existential risk
  4. Students/Parents (Weeks 3-5) — actual customer voice
  5. Policy/Government (Weeks 5-7) — sharper questions after ground truth
  6. Investors (Weeks 5-7) — present informed thesis, get useful feedback

RESEARCH GAPS BEYOND THE 3 ALREADY FLAGGED

# Gap Why it matters
1 Agriculture/agri-tech workforce 42% of workforce, zero coverage. Largest employer in India
2 Informal sector transition 293M e-Shram workers. The largest potential beneficiary population
3 Mental health / learner wellbeing Zero mentions. Kota suicide crisis is directly relevant
4 Disability inclusion 26.8M PwD, 5% reservation mandate. One passing mention
5 Employer L&D budgets Directly affects B2B revenue model. Zero data
6 Tier 2/3 competitor landscape Thousands of local training centers. No mapping
7 Parent decision-making depth WTP by income segment, payment preferences, trust signals. Thin coverage
8 Construction/infrastructure workforce 2M current shortage → 5M in 5 years. Not in framework as a force
9 EdTech crash implications Trust deficit, market reset, B2B pivot. Not captured as context

PRIORITY ACTION PLAN

Tier 1: Fix Before Any External Use (1-2 days)

  1. Fix the 5 critical/high data inconsistencies (PLI jobs, internet access, CSR, NAPS, GER)
  2. Remove or flag the 5 suspicious numbers (35M STEM, 60M EV jobs, 115M jobs, 7M AHP, 4% NPA)
  3. Fix README file names
  4. Add files 08-09 to framework index
  5. Reconcile ops doc number mismatches (staffing, costs, revenue, PMKVY subsidy)

Tier 2: Strengthen Before Expert Interviews (1 week)

  1. Create student/parent interview guide
  2. Create NBFC/lending partner interview guide
  3. Add AI tutor, vernacular economics, clinical practicum, co-location questions to existing guides
  4. Add interview sequencing guide
  5. Fix the 5 too-broad and 2 leading questions
  6. Address cold-start problem in frameworks
  7. Add FIITJEE safeguards to org design, trainer model, and unit economics docs

Tier 3: Build Out Over Coming Weeks

  1. Write competitive landscape framework
  2. Write team & talent strategy framework
  3. Write financial model with sensitivity analysis
  4. Write India Stack integration framework
  5. Create daily operations manual and student lifecycle doc
  6. Create legal/contracts framework
  7. Add safety, insurance, and data privacy docs
  8. Add track-specific trainer ratios (manufacturing 1:15, healthcare 1:15)

OVERALL ASSESSMENT

What's strong: The research breadth is exceptional — 10 files covering demographics through operational costs. The frameworks are intellectually rigorous with genuine strategic insight (trust equation, three markets, flywheel mechanics). The operations docs provide a solid strategic blueprint.

What needs work: Data consistency (11 mismatches, 5 suspect numbers), 4 missing frameworks, 4 missing ops docs, 2 missing expert types, and several load-bearing assumptions with zero validation paths.

Bottom line: The project is approximately 65-70% complete as a strategy package. The missing 30-35% is where execution risk lives — competitive dynamics, financial modeling, legal structure, team strategy, and most importantly, customer and employer validation.