Generative Generalist

Hi, I’m Nandesh.

The generalist who is an AI specialist.

Jack of all trades, king in a few. I was a specialist first, then I kept adding trades until they started compounding. What I do with that: build the revenue operation you keep meaning to hire for, across six specialties, and you leave owning all of it.

Bengaluru  ·  Cornell Chief AI Officer Program  ·  100+ systems shipped

Your revenue pipeline, running end to end

Pick what you sell. The six stages below are what I install, running on a business like yours.

    Inbound, handled
    Hi, is this still available? Need it for a team of 12.02:14
    Yes. For 12 seats that is the growth tier. I have held Thursday 10am for you, does that work?02:14
    Perfect, book it.02:15
    Booked, invite sent.02:15

    This is what it looked like at two in the morning. You were asleep.

    My specialties

    Six specialties.
    Built in this order, on purpose.

    These are the six things I have built over and over, and they stack. You cannot do the AI layer without memory underneath it, and you cannot do memory without foundations. Start anywhere, but this is the order that works.

    Drag through them. Each one is a layer, and each one sits on the one before it.

    See how each one gets built

    Who I work with

    Founders who already have
    something that works.

    I am not much use to a company still looking for its offer. I am very useful to one that has found it and cannot execute fast enough to keep up.

    We will get on well if

    • You are the founder or the operator, and you can decide without a committee.
    • Something already converts. The constraint is capacity, not product.
    • You are two to fifty people, and everyone is doing three jobs.
    • You want to understand the system, not just be handed it.
    • You would rather own the infrastructure than rent it forever.

    You should not hire me if

    • × You want someone to run it forever so you never have to look.
    • × There is no offer yet. Automation multiplies. Multiply zero and you get zero.
    • × Every change needs three approvals and a quarterly planning cycle.
    • × You are shopping purely on hourly rate.
    • × You want AI in a press release more than you want pipeline.

    Sectors I have already shipped systems into. The revenue mechanics rhyme across all of them.

    How I work

    In your accounts, on your data,
    with you in the room.

    No discovery phase that bills for six weeks and produces a document. The first thing I do is build something small that works, so you can judge me on output rather than on a proposal.

    Step 01

    We agree the smallest useful thing

    One call. You tell me what keeps not getting done, I tell you honestly whether I can fix it and roughly what it takes. If I cannot, I say so on that call.

    Step 02

    I build it live, in your accounts

    Not a sandbox, not a mock. Your tools, your data, your credentials, which stay yours the whole time. You watch it take shape rather than waiting for a reveal.

    Step 03

    It ships before the next one is designed

    Real data breaks assumptions in week one that a specification would have protected for a quarter. So the first system goes into production before I plan the second.

    Step 04

    Humans stay on the money moments

    Pricing, an unusual reply, anything irreversible, anything reaching a customer for the first time. The mechanical ninety percent runs itself. The consequential ten percent comes to you.

    Step 05

    The runbook is written as we go

    Not summarised afterwards from memory. Written during the build, in plain language, including the two things that broke and why the obvious approach was wrong.

    Step 06

    I hand over the keys and leave

    Credentials in your vault, repositories under your organisation, workflows in your account. If we never speak again it all keeps running. That is the point.

    The eleven principles underneath this

    The stack

    Tools I have actually shipped with.

    Not a list of things I have read about. Every one of these has been in a system I built, broke, and put back together for a real client or for myself.

    Outbound and enrichment

    • Smartlead
    • Apollo
    • Clay
    • Lemlist
    • Expandi
    • Instantly
    • PhantomBuster
    • Apify
    • Anymailfinder
    • Dropcontact
    • Sales Navigator

    CRM and revenue

    • HubSpot
    • Salesforce
    • GoHighLevel
    • Pipedrive
    • Close
    • Calendly
    • Cal.com
    • Stripe
    • Attio

    AI and models

    • Claude
    • OpenAI
    • Gemini
    • Mistral
    • Ollama
    • Pinecone
    • ElevenLabs
    • Qwen
    • Kimi
    • Whisper
    • Qdrant
    • LiveKit
    • Higgsfield

    Automation and data

    • n8n
    • Supabase
    • PostgreSQL
    • Docker
    • FastAPI
    • Python
    • Prisma
    • Playwright
    • Caddy
    • Hetzner
    • Redis
    • Infisical
    • Cloudflare
    • Restic

    Content and marketing

    • Mailchimp
    • ActiveCampaign
    • Brevo
    • ConvertKit
    • GSAP
    • Remotion
    • Descript
    • Lottie
    • Plausible
    • PostHog

    Comms and operations

    • Slack
    • Notion
    • Jira
    • ClickUp
    • Twilio
    • Gmail
    • Google Sheets
    • WhatsApp Business
    • Telegram
    • Fireflies
    • Zoom
    • Metabase

    Web and interface

    • Next.js
    • React
    • Tailwind
    • TypeScript
    • Framer
    • Figma
    • Vercel

    Your stack is probably in here already. If it is not, it usually has an API, and that is enough.

    Nandesh Goudar
    A bit about me

    I’m the generalist
    who is an AI specialist.

    Bengaluru  ·  Cornell Chief AI Officer Program  ·  Seven years in revenue operations

    I learn the new thing, teach it to whoever is nearby, then build the system that makes it repeat. That has been the pattern since a university classroom, and it is also exactly what I am selling you, except the thing being made repeatable is your revenue.

    Before any of this I spent seven years in commercial and revenue operations. That is the part most AI builders are missing. I have carried a number, so I know which automation actually moves one and which just looks impressive in a demo.

    The generalist half is why I can see the whole pipeline at once. The specialist half is why the AI holds up when it meets real traffic. Most people selling this have one or the other.

    100+
    Systems shipped
    8
    Agents I run daily
    50K+
    Leads generated
    7 yrs
    In revenue operations

    Everything on that list runs right now, most of it on my own server.

    Beyond the systems

    What I bring that is
    not on the invoice.

    Installing the software is the easy half. The half that tends to matter more is what happens in the room while we build it, and what I am reading about at midnight when I am not building yours.

    Thinking about

    Consciousness

    What attention actually is, and whether any of what we are building has a shape like it.

    Thinking about

    Space

    The only industry where the unhappy path gets modelled first. I steal that discipline constantly.

    In the room

    I start the conversations your team is not having.

    Nobody asks me to. Three come up almost every time:

    • What value am I actually bringing to this company?
    • What value am I trying to get this company to bring into my life?
    • How does leadership get built here on purpose, rather than by accident?

    These are not workshop exercises. They land in the middle of building something real, which is the only time anyone answers them honestly.

    Thinking about

    Where AI is going

    Thinking about

    Building what did not exist last month

    Nandesh teaching a full room Most weeks

    In a room, teaching

    Last week

    I built a spherical robot that follows me around.

    It records, it talks back, it takes my calls, and it can drive my terminal for me. Next week it will be something else. You are hiring the person who keeps doing this, not only the one who ships the six specialties.