The Index · IT services

India AI-Native Index · Edition one · Public evidence only

Persistent Systems

Scored on public evidence published between February 2025 and August 2026, reviewed through 30 August 2026, under the Edition 1 locked rule set (ANS-100/P v2.1, 1-floor on every question). Provisional result and evidence annexure sent to the company fourteen days before publication.

59.5Score out of 100
AI IntegratedBand · 41–60 · boundary
AConfidence · 6 of 8 evidence categories
NoneCeilings triggered

Bands are calibrated thresholds on the published scale, not ranks.

What this assessment is, and is not

The Index measures what a company has placed on the public record about its use of AI, within the evidence window, against a published rubric. It does not measure business performance, product quality, safety, customer outcomes, or the adequacy of internal controls. A low score on any question means that, in the public sources listed below, reviewed through the cut-off date, disclosure sufficient to satisfy the published criterion was not located. The company may maintain relevant practices or controls that are not publicly disclosed.

What this assessment is, and is not

The Index measures what a company has placed on the public record about its use of AI, within the evidence window, against a published rubric. It does not measure business performance, product quality, safety, customer outcomes, or the adequacy of internal controls. A low score on any question means that, in the public sources listed below, reviewed through the cut-off date, disclosure sufficient to satisfy the published criterion was not located. The company may maintain relevant practices or controls that are not publicly disclosed.

By dimension

Knowledge & context
50%
Agents & automation
63%
Model & infrastructure
50%
Data & AI security
50%
Organisation
56%
Economics & outcomes
75%
Business-model exposure
65%

Each dimension is the share of available points earned on the questions that could be scored from public evidence. 24 of 35 questions were scoreable; the remainder are outside the public-evidence instrument.

Evidence located

Questions on which the reviewed public sources contained disclosure meeting at least the level-2 anchor of the published rubric:

Level 3 — instances, or a disclosed number linked to a named system

  • AI deployed across functions
  • orchestration or integration platform
  • model strategy or vendor architecture
  • AI leadership and organisation
  • AI attribution to reported economics
  • improvement trail and versioning
  • product that could not exist before AI, with volumes
  • leadership commitment and strategic response

Level 2 — disclosed; not quantified or independently verified

  • named AI platform or knowledge substrate
  • internal AI tooling in use
  • production AI autonomy
  • guardrail and evaluation documentation
  • open or external build surface
  • evaluation and benchmarking practice
  • API or integration surface
  • canonical data platform
  • AI governance posture
  • AI talent or training figures
  • role redesign around AI
  • AI-specific investment
  • AI-native monetisation
  • position on agentic or AI-mediated exposure

Level descriptions are those of the published rubric. The evidence annexure sent to the company lists the specific documents and passages relied on for each question.

Where disclosure meeting the criterion was not located

In the public sources listed below, reviewed through 30 August 2026, we did not locate disclosure sufficient to satisfy the Index's published criterion for the following. The company may maintain relevant practices or controls that are not publicly disclosed.

  • hosting or infrastructure posture

Sources reviewed

"Re(AI)magining the Enterprise," 36th Annual Report (persistent.com PDF, Jul 2026); FY25 BRSR also in pack; Q2–Q4 FY26 + Q1 FY27 pending (Persistent reported Q1 FY27 in Jul 2026; Q3 FY26 and Q4 FY26 analyst-call transcripts archived (persistent.com PDFs); SASVA, iAURA, GenAI Hub, named AI solutions catalogue (Agent Recon, DocIntel, ProcessIntel, Reg(AI)ntel, Pi-OmniKG et al.), Open Source Hub; Broker research desk profile (Jun 2026) documenting Sixth Orbit/SASVA architecture; Microsoft Frontier Firm recognition (partner-grade); Brand Finance fastest-growing IT services brand 2026

Archive references

SourceOriginalArchiveCaptured
SASVA 4.0originalarchived copy2026-09-02
Enterprise AI Transformation / iAURAoriginalarchived copy2026-09-02

Exchange filings and reports relied on:

All sources archived at scoring date; the complete source archive register is available on request via corrections@thecompounding.company. Short quotations, where used, are attributed and used for the purpose of review. Right-of-reply notice sent 10 September 2026 to the Company Secretary, Investor Relations, registered office, corporate communications and legal function. Company response: none received as at publication. No response is characterised as agreement or endorsement.

This assessment is based exclusively on public information available during the stated evidence window. It does not assess undisclosed internal systems or practices. Scores and bands are opinions produced by applying the published rubric to the cited factual record. Not investment advice. The Index is not affiliated with, endorsed by, or sponsored by any assessed company.

Corrections: material factual errors are corrected promptly and logged. See the corrections log and the methodology, including its removal policy.