The signature is the easy part
Collecting an electronic signature is close to a solved problem, and that is precisely why it is the wrong thing to evaluate a system on. If a client later disputes what they agreed to, the signature itself establishes very little: they are rarely denying that they clicked. What they dispute is whether they understood, whether they saw the relevant document, or whether the document they signed is the one being shown to them now. A defensible arrangement answers all three, and none of them is the signature.
Reading evidence, and the ninety per cent rule
The most useful thing a signing flow can capture is evidence of reading. In practice this means tracking how far through each document the signer has actually scrolled and how long they spent,and not releasing the one-time code until a genuine threshold — around ninety per cent of both the agreement and the Investor Charter — has been passed. Clients sometimes find this mildly irritating, which is the point: a flow that can be completed without reading produces a signature that proves nothing, and the irritation lasts a minute while the record lasts for years.
Identity, time and place
A signature is worth more when it is anchored. The anchors that matter are the signer's verified email or phone, the exact timestamp in a stated timezone, the IP address the signature came from, and where possible the city that resolves to. None of these individually proves identity. Together they make a later claim that somebody else signed substantially harder to sustain, and they cost nothing to capture at the moment of signing. Capturing them afterwards is impossible,which is why this is a design decision rather than a feature.
Hash the document, not the template
The detail most implementations miss: record a cryptographic hash of the exact document that was signed, not of the template it came from. Agreements get revised. Disclaimers change. If all you can show is that the client signed 'the MITC', you have to argue about which version. If you can show a SHA-256 hash of the specific file, there is nothing to argue about — either the document being discussed matches the hash or it does not. This single field removes an entire class of dispute.
Put it all on one page, at the end
The practical form of all of this is a certificate page appended to the signed document itself: a certificate identifier, the signer's name, email and phone, the timestamp, the IP and city,the document hash, the seconds spent on each document, when the code was sent, when the document was first opened, and the advisory business's own name and registration number. One page, one PDF, no assembly required at the moment somebody asks. The value of it is entirely in not having to gather it later.
Name your provider, and expect to be asked
Compliance officers ask who performs the signature, and 'we have e-sign' is not an answer. Being able to name a recognised provider — Digio, in our case — answers a question that otherwise turns into a meeting. Similarly, if a digital signature certificate is applied to outgoing documents, its issuer, serial number and validity dates should be recorded for audit rather than just configured and forgotten.
How OptumFlex does this
The agreement and the Investor Charter open in a viewer that tracks reading, and the one-time code does not appear until both are at least ninety per cent read. The signed PDF carries both documents stamped, plus a final certificate page with the identifier, the signer's details, the timestamp in IST, the IP address and city, the SHA-256 hash of the document signed, the seconds spent on each, and the advisory firm's own name and SEBI registration number. Aadhaar e-sign is provided through Digio, and a firm's own digital signature certificate can be applied automatically with its issuer and validity recorded.