Sriram
Murali
I lead engineering for products where money, regulation and scale meet, and I build the tools myself when the market doesn't have them.
I scaled Xoxoday's supply chain from $125M to $1B+ in volume, led Cashfree's move to AI-driven merchant onboarding, and was CTO of Crisp. Now I'm building Beetle, which checks a merchant's compliance from a single URL in about ten seconds.
Scope, not job titles.
What I owned, how big it was, and what changed while I was there.
Payments network
Took a team from below-par delivery metrics to the leading team in its pillar.
The team carried vendor go-live for Asia, Southeast Asia and Europe, and the internal operations dashboard. I made delivery measurable and visible to leadership and managed the team to those numbers: average time to merge a pull request fell from about 72 hours to about 8. I also set the forward roadmap for Command Center, the operations product, to bring operations and newer products into its scope.
Oct 2024 – Sep 2025
Led merchant sign-up, onboarding and the merchant dashboard, and moved onboarding from a day or two to about five minutes.
I ran two squads of ten engineers and partnered with compliance, legal, the CISO and RBI audit teams. The shift to AI-driven onboarding replaced a manual review of one to two days with a decision in minutes, and it showed me what a merchant check costs, who waits on it, and what a regulator expects to see.
Sep 2021 – Oct 2024
Owned the supply chain behind all three Xoxoday products as its volume went from $125M to $1B+.
That supply chain served Plum, Empuls and Compass, and revenue roughly doubled each year alongside the volume. I turned vendor integration into a repeatable product: a standard contract and templated environments cut the time to take a new vendor or market live from about three months to about a week. About 37 vendors were live across Southeast Asia, Europe, MENA and South America when I left.
I also took ownership of audit readiness. When the first external ITGC audit failed on operational procedures, I rebuilt the process and its evidence, worked directly with the auditors, and made every deploy on the core dashboard traceable to the commit behind it.
2014 – 2020 · formerly Tenreads
Built a news-summarisation engine that reached 80–85% accuracy years before LLMs.
I was there from the first day, owned the technology end to end, and built the engineering team. Two things came out of it.
The engine was the product: timelines generated automatically from the news flow and served through an API that partners could embed.
Bloomberg Quint, our first design partner, embedded those timelines in its story pages for a 30% share of ad revenue. It moved the company's north star from session time to revenue.
When the company shut down in 2020 I kept the product running myself: about half a million stories on a server that costs $4 a month.
Beetle: merchant compliance in about ten seconds.
Payment gateways check a merchant at onboarding and again when volume crosses a threshold. Nothing triggers when the merchant's website changes. The checks that do exist take five to thirty minutes at best.
Beetle takes one URL and returns what the merchant sells, the likely category codes with confidence, whether the policy pages are in place, and the evidence behind each call. It then re-checks daily and sends a webhook if anything material changes.
I came to this from the business side as much as the engineering side. Leading onboarding at Cashfree meant sitting between compliance, legal, the CISO and the engineers, so I know what a check costs, who waits on it and what the regulator will ask to see. Beetle is that experience as a product: priced per check, built for the teams that carry the compliance budget, with the evidence behind every decision kept.
Three habits that show up everywhere I've been.
Let the platform mature with the product.
Build the foundations as reusable components from the start (contracts, deployment, audit trails), so that by the time the product matures they have become the platform. Xoxoday's integration framework and Beetle both grew this way.
Lean on purpose.
Half a million stories run on a $4 server. Beetle runs on a couple of small ones. Tight limits force a clear view of what is actually load-bearing.
The org chart is architecture.
Clear lanes and a steady rhythm decide whether a dozen engineers move as one team or as a dozen. I do the hands-on work myself and spend the rest of my time on the team's growth.
Notes from the work.
The longer thinking behind the work, the decisions and the trade-offs, lives on the blog.
Read the blogWhat others wrote about Crisp.
“Tenreads' Timelines currently operate at an accuracy of 80–85 percent, which is extremely high for an AI system.”
Read →“Tenreads aims to stem the filter failure problem and help people find the ten best stories everyday on the things they care about.”
Read →“Whatever other apps are doing with dozens of writers and editors, we are doing as four engineers.”
Read →Let's talk about what you're building.
I'm interested in two kinds of conversation: engineering leadership roles (CTO, VP or Head of Engineering) at companies where compliance, payments or large-scale operations are part of the product, and investors or partners who want to look at Beetle.
Write to me directly.
+1 408.569.8429 US