Here is the full article, verbatim, for copy-paste:
Finarkein AA Report: When a Vendor’s Own Data Becomes Industry Analysis
SEO Metadata
- Focus Keyphrase: Finarkein Account Aggregator report
- Title: Finarkein Account Aggregator Report: When a Vendor’s Own Data Becomes Industry Analysis
- Meta Description: Finarkein’s AA ecosystem report at GFF 2026 proposes a 23-factor AI framework. But the data comes from its own infrastructure — is this analysis or marketing?
- URL Slug: finarkein-account-aggregator-report
- Tags: Finarkein, Account Aggregator, AA ecosystem, Sahamati, GFF 2026, open finance, digital public infrastructure, AI in lending, data reliability, consent-based data sharing, NBFC-AA, RBI, financial data, credit underwriting, fintech India
- Excerpt: Finarkein’s GFF 2026 report on India’s AA ecosystem proposes a 23-factor framework for AI accountability. The data is drawn from its own platform — raising questions about what is industry analysis and what is vendor positioning.
- Categories: Fintech, Digital Public Infrastructure, AI in Finance
- Subcategories: Account Aggregator, Open Finance, AI Governance, Credit Infrastructure
Opening
On 11 September 2026, Pune-based Finarkein released a report titled “On Systems of Intelligence: Evidence and Obligations in the AA Ecosystem” at Global Fintech Fest (GFF) 2026. The Finarkein Account Aggregator report analyses transaction data observed on Finarkein-provided infrastructure between December 2025 and August 2026, covering 95 institutions and 111.1 million consent requests.
The report arrives at a pivotal moment. RBI recognised Sahamati as the Self-Regulatory Organisation for the AA ecosystem in June 2026, and the ecosystem has crossed 493 million fulfilled consents. AA-enabled lending has reached meaningful scale. Finarkein’s pitch is that the conversation must shift from volume to reliability and intelligence.
What the press release does not foreground is that Finarkein is both the author and the subject of this report. The data comes from Finarkein’s own platform. The framework it proposes would govern the very products Finarkein sells. And Finarkein is the official “Account Aggregator and AI Partner” of GFF 2026 — a paid sponsorship position, not a neutral research designation. This article separates the genuinely valuable signals from the vendor positioning.
Why Now: The Timing Logic
A newly regulated ecosystem
The RBI recognised Sahamati as the SRO for the AA ecosystem on 5 June 2026. That move formalised governance over a network of 17 operational Account Aggregators, 176 Financial Information Providers, and over 1,000 Financial Information Users. The SRO framework tasks Sahamati with developing operational and technical standards, facilitating dispute resolution, and coordinating ecosystem participants.
The Finarkein Account Aggregator report lands squarely in the window where this governance is being defined. Its call for service-level standards, error-rate transparency, and accountability frameworks aligns with what an SRO would need to operationalise. Whether that alignment is coincidental or strategic is worth noting.
Ecosystem scale, but whose numbers?
Sahamati’s own dashboard, updated through July 2026, shows 493 million cumulative consents fulfilled, 326 million linked accounts, and 1,076 FIUs. The press release cites 45 crore (450 million) fulfilled consents and 1,020 FIUs — figures that are slightly stale but directionally consistent with Sahamati’s April 2026 data. Sahamati’s own Credit Reimagined report estimated AA-facilitated lending at a comparable level for H1 FY26.
What matters here is the overlap, not the difference. The headline ecosystem numbers in the Finarkein report are not Finarkein’s own findings. They are Sahamati’s publicly published statistics, restated within a Finarkein-branded document.
The company behind the report
Finarkein was founded in 2019 by Nikhil Kurhe and Dheeraj Kumar. It employs 99 people, operates from Pune, and has raised $6.25 million across multiple rounds. Its investors include Nexus Venture Partners, IIFL Fintech Fund, Info Edge Ventures, Eximius Ventures, and the DSP Group Family Office. Its most recent round was a $1.5 million extended pre-Series A in October 2025.
The company positions itself as a “System of Intelligence for financial institutions” — a term borrowed from broader enterprise software discourse about AI layers sitting above databases. Finarkein’s product portfolio includes multi-AA orchestration, data analytics, and AI-powered decisioning across underwriting, collections, and fraud detection. GFF 2026 listed Finarkein as its official “Account Aggregator and AI Partner and Exhibitor” — a sponsorship role, not an independent research designation.
The Competitive Picture
Multi-AA orchestration is a crowded field
Finarkein is one of several technology service providers (TSPs) in the AA ecosystem offering multi-AA orchestration. The competition is substantial.
Setu, backed by Pine Labs, offers a multi-AA gateway with intelligent AA-FIP routing across 100+ FIPs and 5+ AAs. MoneyOne brands itself “India’s Largest AA TSP” with smart AA routing across 120+ FIPs. Ignosis provides real-time multi-AA orchestration with live FIP and AA health monitoring, and counts Moneyview among its customers.
Perfios, a larger and more established player, offers its FIU++ product with 50+ FIPs integrated and pre-built connections to Sahamati’s Central Registry. Signzy offers an AA API with bank statement analytics. OneMoney, India’s first RBI-licensed AA itself, also provides TSP middleware through its FinPro and FinShare products.
Where Finarkein stands
What separates Finarkein is not the multi-AA orchestration model, which is table stakes in this category. It is the combination of orchestration with an AI and analytics layer that turns raw financial data into “judgments” — credit assessments, risk scores, collection strategies. Competitors like Setu and MoneyOne focus primarily on routing and data fetch reliability. Perfios competes more directly on analytics but operates across a broader product surface.
The Finarkein Account Aggregator report is, in part, a positioning document. By defining the problem (data reliability, AI accountability) and proposing the framework (23 factors), Finarkein is shaping the terms on which its own products will be evaluated. That does not make the framework wrong, but it makes the conflict of interest worth stating plainly.
Sahamati already publishes this data
Sahamati runs a public ecosystem dashboard with monthly statistics on accounts linked, consents fulfilled, and data shares. It also publishes detailed reports — the Credit Reimagined series for lending, the Data Unlocked report for capital markets and insurance. These are ecosystem-wide, drawing from all participating entities.
The Finarkein report draws from 95 institutions on Finarkein’s own infrastructure. That is a meaningful sample, but it is not the ecosystem. The distinction matters when headlines cite “the AA ecosystem” and the data underneath is from one vendor’s platform.
What the Data Shows
The genuinely interesting findings
Three findings in the Finarkein Account Aggregator report add real value beyond what Sahamati publishes.
First, deposit data accounts for roughly 91 per cent of requests. Mutual funds follow at 5.33 per cent, equities at 2.15 per cent, and ETFs at 1.09 per cent. This concentration in core banking data is not a new observation, but having a specific breakdown from a live platform is useful. It highlights how far the ecosystem remains from serving wealth and insurance use cases.
Second, 52 of 74 FIPs were accessible through more than one AA, and multi-routed FIPs accounted for 99.6 per cent of sessions. This is a genuine infrastructure insight. Multi-routing means a lender can reach the same bank through different AA paths, creating a built-in redundancy mechanism. It also means performance can be compared across routes — a prerequisite for the SLA accountability the report calls for.
Third, the performance data is specific and sobering. Across five AAs, 3,898 error responses were recorded between March and August 2026. Four of five AAs typically responded under 0.2 seconds. The slowest 1 per cent of calls at the fifth AA reached 1.4 seconds for data requests and 2.3 seconds for data fetches.
The report notes that no adopted service-level standard exists for these machine-to-machine timings. That gap is real and worth flagging.
What the data does not show
The report’s 45.4 per cent consent approval rate and 75.9 per cent data fetch success rate are headline figures. What the press release does not address is the other side of both numbers. Why did customers not approve the 54.6 per cent of consent requests? And what about the 24.1 per cent of data fetches that failed?
Are the unapproved consents a customer-experience problem (too many requests, confusing interfaces) or a fraud-prevention signal (customers rejecting suspicious access attempts)? Are the failed fetches caused by FIP downtime, AA routing errors, or something in Finarkein’s own orchestration layer? Without answers, it is hard to judge whether AA infrastructure is ready for consequential decisions.
The 23-factor framework
Finarkein’s report proposes a 23-factor framework across five areas — source, method, output, system, and disclosure — to assess the quality and reliability of data and intelligence products built on AA infrastructure. The framework targets a real problem. As technology providers move from delivering raw data to delivering “judgments” — credit scores, risk assessments, collection strategies — the accountability gap widens.
That framework has not been independently adopted, validated, or peer-reviewed. No regulator has endorsed it. No industry body beyond Finarkein has committed to it. It is a proposal from a vendor that sells the very products the framework would govern.
Its value depends on whether Sahamati, the RBI, or the broader ecosystem picks it up. It also depends on whether Finarkein is willing to have its own products evaluated against it transparently.
What’s New vs. What’s Repackaged
Genuinely new
The 23-factor framework. Proposing a structured way to evaluate AI and intelligence products built on AA data is a genuine contribution. Whether it gains adoption is an open question, but the framing — source, method, output, system, disclosure — is a reasonable starting point for a conversation the ecosystem needs to have.
Cross-AA performance benchmarks. The error counts and latency figures across five AAs are the kind of operational data that no one else has published publicly in this level of detail. If Finarkein continues to report these quarterly, it would create a useful accountability mechanism.
The multi-routing insight. That 52 of 74 FIPs are reachable through multiple AAs, accounting for 99.6 per cent of sessions, is a concrete data point about ecosystem redundancy. It has implications for reliability design that go beyond what Sahamati’s aggregate dashboard shows.
Repackaged
Ecosystem statistics. The headline numbers — 17 AAs, 176 FIPs, 1,020 FIUs, 45 crore consents — are Sahamati’s publicly published figures, restated in a Finarkein-branded document. They are accurate but not original to this report.
The “Systems of Intelligence” framing. The concept of a “system of intelligence” sitting above a “system of record” comes from enterprise software discourse, popularised by analysts and venture firms. Finarkein applies it to AA infrastructure competently, but the framing is borrowed, not invented.
The call for data quality and reliability. Every ecosystem participant — Sahamati, RBI, lenders, and other TSPs — has been saying this. Finarkein’s specific data points are new; the overarching message is not.
Unclear
How representative is Finarkein’s platform data? The report covers 95 institutions; the ecosystem has over 1,000. Finarkein’s 95 may skew toward lenders and fintechs that chose Finarkein specifically, which could bias the performance and consent-rate findings.
Whether the framework will be open. Finarkein has not stated whether the 23-factor framework will be open-sourced, contributed to Sahamati, or kept as a proprietary differentiator. That decision will determine whether it becomes an industry standard or a marketing tool.
The Question That Wasn’t Answered
Who is auditing the auditor?
Finarkein is a technology provider selling AI-powered decisioning products. It is also the entity publishing a report that proposes standards for evaluating AI-powered decisioning products. Who evaluates Finarkein’s own products against the 23-factor framework? The press release does not address this structural conflict. A framework proposed by a vendor, applied to the vendor’s own category, without independent oversight, is a marketing asset until proven otherwise.
What happens to the 54.6 per cent who did not approve?
The consent approval rate of 45.4 per cent means more than half of all consent requests were rejected or ignored. For an ecosystem built on user consent, that is a striking figure. Is it a sign of healthy user agency, or evidence that the consent experience is broken? The report does not analyse the rejection side of the funnel.
Is the performance data flattering Finarkein’s own routing?
The error and latency figures are measured across AAs, but the data flows through Finarkein’s orchestration layer. If Finarkein’s routing logic selects the best-performing AA for each request, the performance numbers reflect Finarkein’s optimisation, not the raw AA infrastructure. That would make the figures look better than an unoptimised baseline — useful for Finarkein’s sales pitch, but potentially misleading as an ecosystem health indicator.
What This Means for You
For lending and FIU decision-makers
The multi-routing and performance data in the Finarkein Account Aggregator report are worth reading for their operational detail. If you are evaluating AA infrastructure, the fact that four of five AAs respond under 0.2 seconds is reassuring. The fact that the slowest 1 per cent at one AA reaches 2.3 seconds is a risk to factor into your SLA expectations. Ask your TSP which AAs they route through, and whether they have fallback logic for failures.
For regulators and Sahamati
The 23-factor framework is a useful conversation starter, but it should not become a de facto standard simply because Finarkein published it first. If the SRO adopts a framework for evaluating intelligence products, it should be developed through a multi-stakeholder process, not imported from a single vendor. The SLA gap the report identifies is real and worth prioritising.
For investors and fintech watchers
Finarkein is a 99-person, pre-Series A startup with $6.25 million in funding. Publishing an “ecosystem report” at the industry’s largest event is a classic category-leadership play — positioning the company as the thought leader in a space it wants to dominate. The report’s quality is genuine in parts, but the strategy is unmistakable. Evaluate the data on its merits; evaluate the company on its execution, not its report titles.
Editor’s Note
This article draws on the Finarkein press release issued via Genesis BCW on 11 September 2026, supplemented by independent web research. Company-claimed information includes all statistics sourced from “Finarkein-provided infrastructure” — 95 institutions, 111.1 million consent requests, 45.4 per cent approval rate, 75.9 per cent fetch success rate, 3,898 error responses, and the 23-factor framework.
Independently verified information includes Sahamati’s ecosystem dashboard data (493M+ consents fulfilled, 176 FIPs, 1,076 FIUs as of June/July 2026). The RBI’s SRO recognition of Sahamati, Finarkein’s funding history, and its GFF 2026 sponsorship role are also verified. So is the competitive landscape of multi-AA TSPs — Setu, MoneyOne, Ignosis, Perfios, Signzy, and OneMoney. The ecosystem-wide statistics cited in the press release (17 AAs, 176 FIPs, 1,020 FIUs) match Sahamati’s April 2026 data, not Finarkein’s own findings. We could not independently verify the 23-factor framework’s adoption, validation, or peer review status.