San Francisco Bay Area

Vivek Raghunathan

Engineer · Founder · Angel Investor

20 years building systems where failure isn't an option.

↓

Introduction

I’ve spent about 20 years building systems where failure isn’t an option — from device drivers on educational tablets to AI platforms inside large financial institutions.

I was employee #2 at Kno, where we grew engineering to 80 and were acquired by Intel. At Intel I led the backend for an education platform serving 100K+ students. Founding-team engineering at Togg followed, and for the past eight years I’ve been CTO at Decision Minds, putting data and AI to work inside financial institutions while co-founding Success4 and running it to 25 customers.

Today I’m founder and CEO of Adalma AI: agentic AI for credit unions and community banks, built so the work stands up to an examiner. On the side, I angel invest in infrastructure and AI, and I build things — trading systems, agents, apps — mostly because I can’t help it.

Career

Two decades, one throughline: systems that can't fail.

  1. 2026 — Present

    Adalma AI logo

    Adalma AI

    Founder & CEO

    • Compliance-first AI for credit unions and community banks — compliance enforced before the action, not caught after
    • Cernio takes financial-crime work from alert to attested, filing-ready SAR, with automated SAR and CTR e-filing to FinCEN. Cura recovers past-due accounts with TCPA, Reg F and FDCPA enforced in-line. Tessra policy-gates and logs every model call
    • First customer in production since August 2026, forward deployed inside their stack
    • Tamper-evident audit trails an examiner can verify, running in your environment. See it live at adalma.ai/proof

    CernioCuraSentraTessra

  2. 2018 — Present

    Decision Minds logo

    Decision Minds

    CTO & Client Partner

    • Putting data and AI to work inside large financial institutions for eight years
    • Replaced the company's entire HRMS, CRM, recruitment, and expenses suite with homegrown platforms — Keta and a purpose-built CRM — sharing one common data platform and saving close to $60K a year in SaaS spend
    • Keta covers the full employee lifecycle: payroll, expenses, approval flows, FDE programs, e-signatures (replacing Eversign/Xodo), and SharePoint integration
    • Employees submit time and expenses straight from WhatsApp — with a Teams bot alongside for the rest of the day-to-day
    • The CRM runs the full sales pipeline with an Outlook plugin, a Chrome extension for LinkedIn, and enrichment from Apollo and other data sources
    • Roadmap: agentic automation woven through the entire suite

  3. 2018 — 2026

    Success4, Inc logo

    Success4, Inc

    Founding CTO

    • Co-founded Success4 and ran it for eight years alongside the Decision Minds CTO seat — B2B SaaS for customer success, 25 customers at its peak
    • All things engineering, in service of customer success

  4. 2016 — 2018

    Togg

    Founding Team — Head of Engineering

    • Built the engineering org from the founding team out

  5. 2013 — 2016

    Intel logo

    Intel

    Backend Engineering Manager & Tech Lead

    • Led backend for Intel Education Study platform serving 100K+ students
    • Content catalog spanning 1M+ titles and 100M+ data points
    • Led an Intel Education Service Corps assignment to Rwanda — international team excellence award from the Rwanda Education Ministry

  6. 2009 — 2013

    Kno, Inc. (acquired by Intel)

    Senior Software Engineer & Software Lead

    • #2 non-founding employee — helped engineering grow to 80 and through acquisition by Intel
    • Embedded Linux, device drivers, kernel work on next-gen tablet hardware

  7. 2008 — 2009

    Juniper Networks logo

    Juniper Networks

    DVT Engineer

    • Design verification and testing of switch control and interface boards

Anna University Chennai, B.E. ECE → Florida International University, M.S. EE

Angel Investing

Backing builders in infrastructure and AI.

Angel investing since 2019 — infrastructure, AI, and the occasional conviction bet on a founder. Member of The Chennai Angels since 2020, bridging the US and Indian startup ecosystems.

13+investments
2ecosystems (US · India)
3via The Chennai Angels

Side Projects

I build things because I can't help it.

Perspective

My take on the AI wave

Every platform shift rewards the people who ship through it, not the ones who spectate. I’ve watched this movie before — mobile, cloud — but this one is different in one specific way: the marginal cost of trying an idea has collapsed to nearly zero.

Three beliefs drive how I spend my time:

Agents are the new backend. The interesting systems being built right now are not chatbots — they’re pipelines of small, verifiable, autonomous steps with gates and audit trails. That’s how I build my trading systems, and it’s how we build compliance systems at Adalma.

Regulated industries are the real frontier. Anyone can demo AI. Making it stand up to an examiner — policy-gated, logged, tamper-evident — is where the durable value is. The institutions that need AI most are the ones that can afford it least, and closing that gap is a business, not charity.

Taste survives automation. When everyone can generate everything, the scarce skill is knowing what’s worth building and when to say no. Twenty years of shipping has taught me that judgment compounds faster than code.

The next decade

Three people frame how I think the industry shapes up from here. Satya Nadella has argued that the software stack collapses — SaaS thins out into agent-orchestrated workflows, and the business logic migrates to whoever runs the agents. Elon Musk keeps pulling the conversation down to physics: intelligence per watt, compute as a manufacturing problem, and humanoid robots as the largest product category ever. And Nikesh Arora asks the question every enterprise will be forced to answer — not “how do we adopt agents” but “how do we discover, govern, and stop them when we need to.” Stack those together and the next decade looks like agents doing the work, physical AI doing the labor, and governance deciding who gets to sell any of it to serious institutions.

Energy becomes the commodity of the era. Every one of those futures is gated on electrons. Data centers are already negotiating for gigawatts the way they once negotiated for bandwidth, and the winners of the next decade will be the ones who solved generation — next-generation nuclear, solar plus storage, geothermal — and the quieter constraint underneath it all: water. Cooling, agriculture, and communities compete for the same acre-feet, so water recycling and desalination stop being environmental footnotes and become infrastructure businesses. Watch where the energy and water deals happen; that’s where the compute will be.

Robots come home. The first wave of physical AI is industrial, but the one that matters socially is domestic — machines that fold the laundry, watch the stove, and above all help seniors live independently for another decade. Care is the most under-served labor market on earth, and the demographics are unforgiving. A robot that can steady a hand, fetch medication, and call for help is not science fiction anymore; it’s a product roadmap. Having spent years thinking about the people fraud targets first — seniors on fixed incomes — I think dignity-preserving eldercare is one of the most consequential things this wave will ship.

And the one I’m most excited about: agrotech. Specifically, replacing harsh chemical pesticides with UV-based pest control — autonomous robots sweeping fields at night, using calibrated UV light to kill pests and mildew with no chemical residue, no runoff, and no resistance treadmill. Companies like TRIC Robotics are already running tractor-scale autonomous UV robots on California strawberry farms, and the early results are real: chemical-free control that pests can’t develop resistance to. Pair that with precision irrigation and soil-level sensing, and you get food grown with less poison and less water at the same yield. Agriculture never gets the headlines, but it’s where AI, robotics, energy, and water all intersect — and it might be the wave’s most quietly transformative act.

Writing & Feeds

Thinking out loud.

Day 2: options sleeve made its first real trades (MU, SPCX, AMZN, AXON calls), a bug blocking every order since day one is fixed. AXON hit its stop loss and closed for a loss instead of riding it. AI council rejected a SNDK trade over a data integrity issue. @AlpacaHQ @lablabai

2026-09-01 on X →

Day 1 of my Alpaca hackathon agent trading autonomously in the cloud. It placed zero option orders. Four gates blocked it, each only visible after fixing the one before: 1. sleeve off via an inherited flag 2. no tick inside its entry window 3. greeks 400d, 100 symbol cap 4. limit price needed 2 decimals Every run logged healthy. @AlpacaHQ @lablabai

2026-08-31 on X →

Day 1 of the @AlpacaHQ x @lablabai AI Trading Agents Hackathon. Team: Refusal Rails. Solo entry. The pitch: trading agents get judged on the orders they place. Mine is built around the orders it declines. 22 named risk gates sit between every decision and every order, and every gate that fires is logged with the reason. Fresh paper account. Zero trades so far, on purpose: the market closes in half an hour and my agent refuses positions it can't monitor into a weekend. It starts trading Monday, when it can pass its own gates. #buildinpublic

2026-08-28 on X →

Day-to-day: @nkrvivek on X

Beyond Work