How Deepank Bhandari Is Rewriting India’s Investment Banking Future with S45 and AI
As India’s capital markets grow rapidly, AI-native investment bank S45 is transforming IPOs and M&A with faster execution, stronger compliance, and smarter deal-making.
India’s investment banking sector is growing rapidly, with the market expected to expand at a CAGR of 8–10% over the next few years, driven by rising IPO activity and increasing institutional participation. In 2025, India recorded 367 IPOs, highlighting strong market momentum alongside stricter compliance standards.
In this interview, Deepank Bhandari, Founder of S45, India’s first AI-native investment bank, shares how the company is using AI to transform IPO readiness, improve disclosure accuracy, and bring greater speed and efficiency to capital market transactions.
StartupTalky: S45 describes itself as India's first AI-native investment bank. Traditional investment banking is built on trust, relationships, and judgment. How do you convince a founder preparing for an IPO that an AI-augmented platform can outperform the relationship-driven human model they are used to?
Deepank Bhandari: We do not convince founders by saying AI replaces trust. That would be the wrong framing. Investment banking is still built on trust, judgment, and accountability. What has changed is the execution layer underneath that trust.
A founder preparing for an IPO is not just hiring a relationship. What they actually need is someone who can bring precision, speed, and disclosure quality to a genuinely complex process, without the business taking a backseat while the transaction runs. When everything depends on human bandwidth, you naturally get delays, inconsistency, and things slipping through the cracks.
At S45, AI does not replace banker judgment. What it does is make the process far more reliable. Our systems handle the document-heavy groundwork: they read filings, find inconsistencies, organise disclosures, identify investors, highlight missing information and keep a clear record. Humans then make the decisions that need market knowledge, understanding of regulations and good judgment.
StartupTalky: S45 can reportedly assess an IPO-ready company in 90 minutes and prepare a DRHP in 7 days against the industry norm of weeks or months. What are the specific AI workflows enabling this compression and where does human judgment remain irreplaceable?
Deepank Bhandari: The speed comes from rebuilding how the work actually gets done, not from layering AI on top of an unchanged process. A traditional IPO has endless back-and-forth loops, such as gathering company information, going through financials, reviewing corporate documents, tracing related-party transactions, checking litigations, drafting business sections, reconciling disclosures, and keeping founders, auditors, lawyers, and bankers all moving together. Almost all of it is document-heavy and highly repetitive, and a single error in any one layer can create problems downstream.
Our AI workflows convert unstructured documents into structured outputs, extracting information from financial statements, board documents, ROC filings, cap tables, and contracts. We also use workflow agents for readiness assessment, diligence checklists, investor mapping, and DRHP drafting.
That said, there are things we would never hand over to a system. Whether a company is truly ready to go public, how to frame its equity story, and how to handle a regulatory grey area. These still need experience, context, and judgment.
StartupTalky: India had 367 IPOs in 2025, yet three out of four NSE applications were rejected in the past two years, often on disclosure grounds. How specifically is S45's platform reducing disclosure errors and what is your quality assurance model when regulatory compliance is at stake?
Deepank Bhandari: Disclosure errors usually happen because information is scattered. One fact in the financial statements, another in a board resolution, another in a management conversation. In a manual process, reconciliation depends on people remembering every dependency.
Our platform is built around source traceability. Every output must map back to a source document. Our quality assurance model has three layers: automated checks for missing data, contradictions, and disclosure gaps; domain reviewers validating material sections; and senior bankers applying judgment before anything moves forward. In regulated workflows, AI should never be a black box. It should be an evidence engine. The objective is not only faster drafting, but it is also higher-confidence disclosure.
StartupTalky: S45 has declined 75 of 80 evaluated companies. What is the most common reason a company fails S45's readiness assessment and how do you communicate that rejection constructively to the founder?
Deepank Bhandari: The most common reason is not a lack of ambition. The issue is readiness. Many companies have strong businesses but weak governance, immature financial controls, promoter structures needing cleanup, or an equity story not yet strong enough for institutional investors.
We try to make that conversation as constructive as possible. We sit down and walk through what we actually saw: where the business is today, what is holding it back, and what fixing that realistically takes. Honestly, for a lot of founders, the most valuable thing we can tell them is to slow down. Build for 12 to 18 more months, clean up the house, and come back stronger. A bad IPO is far worse than a delayed one.
StartupTalky: You scaled one of your ventures to INR 500 Cr in revenue before founding S45. What did that experience teach you about the gap between strong fundamentals and access to the right capital and how is that personal insight shaping S45's design?
Deepank Bhandari: That experience made the problem very personal. My co-founder had built a fundamentally strong business: real revenue, real customers, real operating depth. But strong fundamentals alone do not automatically translate into the right capital outcome. That gap is what stayed with us.
In India, many serious founders build strong companies but lack the institutional machinery to tell their story correctly or reach the right investors. That is one of the core reasons we built S45. We come in early and do the work properly by assessing readiness, identifying gaps, shaping the equity story, and executing with precision. A founder who has built something genuinely strong should not lose out simply because the capital markets process is opaque or dependent on knowing the right people.
StartupTalky: S45 is now expanding from IPOs into M&A. India has thousands of companies in the $12 to 36M revenue range where consolidation is inevitable. How are you positioning S45 to capture this wave and what variables in M&A most resist AI automation?
Deepank Bhandari: What we are seeing in the Indian mid-market is a lot of founders reaching a natural inflection point. They have put in ten to fifteen years and built real businesses, but growth from here needs something different. Maybe it is a succession question, maybe margins are getting tighter, maybe they have hit a distribution wall. And separately, buyers and sponsors want exposure to well-run Indian businesses but struggle to find and evaluate them at scale.
S45 is building the infrastructure to identify, assess, prepare, and match these companies. But M&A has variables that resist automation. Founder psychology, negotiation dynamics, family alignment, and timing are deeply human. The actual deal still requires trust and judgment.
StartupTalky: You have argued publicly that "raising capital is not the endgame enduring companies are built across cycles." For a founder currently being pressured by investors to pursue a quick liquidity event, what is your advice?
Deepank Bhandari: My advice is simple: do not confuse liquidity with destiny. If the company still has compounding potential, be careful about optimising for short-term pressure. Capital markets reward companies that survive cycles, not just capture windows.
Founders should ask three questions. Is the business becoming stronger with time? Is the market opportunity still expanding? Will this transaction improve the company’s long-term position or only solve a short-term stakeholder need?
If the answer is the latter, pause.
Enduring companies are built through discipline, not urgency. The best founders use capital as a tool, not as validation.
StartupTalky: S45 is a startup disrupting an established, relationship-dependent industry. What has been the most underestimated source of resistance in the market, and how have you navigated it?
Deepank Bhandari: The most underestimated resistance is not technology skepticism. It is the industry's comfort with opacity. Traditional investment banking has manual processes that survive because they are invisible to the client. When you introduce systems, audit trails, and measurable quality, you make the process transparent. That changes expectations.
We are clear that S45 is not anti-human and not anti-relationship. We are rebuilding the operating layer of investment banking. But execution should not depend on memory, scattered documents, or endless manual follow-ups. The future investment bank will be a smaller, sharper institution were systems own execution and humans' own judgment. That is what we are building at S45.
Preeti Karna
