● Shruti Dyundi · Demand generation, pipeline, GTM systems

Most growth teams have amnesia.

Every quarter, B2B teams re-run experiments they already ran, re-learn lessons they already paid for, and rebuild strategy that retired with someone's laptop. Fifteen years across media, brand, growth, and B2B SaaS demand generation, and the pattern never changes: organizations generate more knowledge than they keep.

So that's what I build. Revenue systems that remember: pipeline engines, revenue operating models, and lately, software.

Director of Global Demand Generation at Tigera · Building Teamless

Start a conversation Read the field notes

What I build → Pipeline engines· Revenue operating models· Decision & learning systems· AI-enabled workflows· Teamless

Problems I know how to solve Triage order

Three problems I keep getting handed.

Pipeline you can't predict.

"We need more leads" is almost never the diagnosis. I build engines: ICP definitions with teeth, intent and signal orchestration, inbound, outbound, and SDR capacity models, and coverage math that survives a board meeting. Quality, velocity, and predictability over top-of-funnel volume, every time.

A revenue org that isn't one.

Demand gen, SDRs, sales, product marketing, and RevOps each optimizing their own dashboard is five teams losing politely. I align them under one operating model: shared definitions, shared numbers, shared consequences. One funnel dictionary. One forecast both leaders sign.

Decisions that don't compound.

Attribution nobody trusts, experiments nobody owns, forecasts built on definitions nobody shares. I instrument the decision layer: attribution held as evidence, experimentation with named owners and stakes, and reporting that changes what happens next quarter instead of describing the last one.

Selected operating work Claims come with receipts

Claims come with receipts.

The evidence is in the systems, decisions, and operating artifacts. Three problems I have owned end to end, and what it took to run them.

Case 01 · Pipeline engine

Building a pipeline you can forecast.

The ask is usually "more leads." The problem is usually definitions. I start where forecasts actually break: who we sell to, what qualified means, and which buying signals justify SDR time. Then I connect the pieces most teams run separately: ICP and qualification standards, intent and signal routing, inbound and outbound capacity, and coverage math by motion. The result is a pipeline forecast sales leadership can interrogate and trust, because the definitions and assumptions are shared.

Artifacts · ICP definition · qualification standards · signal thresholds · SDR capacity model · coverage math by motion · forecast cadence

Case 02 · One revenue model

Making five functions run one operating model.

Demand gen, SDRs, sales, product marketing, and RevOps often optimize different definitions and dashboards. I have built and operated versions of a unified revenue model: shared funnel definitions, explicit decision rights, common qualification standards, and one view of pipeline performance. Alignment stops being a recurring meeting topic and becomes part of how the system runs.

Artifacts · funnel dictionary · stage definitions · qualification standards · decision rights · one view of pipeline

Case 03 · Organizational memory

Making the learning compound.

Every campaign generates knowledge: what the segment ignored, which sequence outperformed, why the numbers moved. Almost none of it survives the quarter. I structure the loop so it does: experiments with named owners, decisions recorded with their reasoning, and retrospectives that write directly into the next brief. Institutional memory, run first as an operating discipline, and eventually as software.

Artifacts · experiment registry · decision records · retros that feed the next brief · attribution held as evidence

this one refused to stay a process. it became Teamless.

Selected operating principles Written down so I can be held to them

Quality over volume. Volume flatters dashboards. Quality closes.

Blended metrics hide sins. Report the split. If the split embarrasses a channel, the report is working.

Systems over heroics. If it worked because one person pulled a heroic week, it didn't work. It borrowed.

Strategy should leave a trace. If nobody can reconstruct why a decision was made, the organization cannot learn from the result.

Attribution is evidence, not truth. Useful for arguing better. Dangerous for declaring winners.

AI systems Labeled with today's status

I don't "use AI." I run systems.

The difference: a system has inputs, checks, a place where the model works, and a place where judgment stays human. Three of mine, each labeled with what it actually is today. One of them barely uses a model at all, which is the point.

System 01 In regular use

The daily content engine.

An Airtable staging hub and a Claude skill support my LinkedIn content workflow, checking each new draft against everything already published and flagging repetition before review. The final judgment, edits, engagement, and publishing remain mine.

MODEL · drafting, archive checks, repetition flags
MINE · judgment, edits, engagement, publishing

System 02 In use · manual

The decision record.

Every significant call, a campaign, an experiment, a budget shift, closes with its decisions, results, and surprises captured in a structured record, and the next brief begins by reviewing that history. In use today as a manual discipline; the productized retrieval and learning layer is part of Teamless V2.

MODEL · not yet. productized retrieval is Teamless V2 work
MINE · what gets recorded, what it means, what changes next

System 03 Active build

The Teamless build pipeline.

V2 product iterations and marketing assets run through model-in-the-loop pipelines, with automated visual checks on assets before they ship. AI expands what I can explore and test. It doesn't inherit the product judgment.

MODEL · generation, tests, visual QA
MINE · product decisions, taste, the roadmap

Where I've done the work 2011 → present

2011 → 2016 GroupM

Media strategy and measurement for large consumer brands. Where I learned that reach is a commodity, judgment is not, and every attribution model has opinions.

2016 → 2019 Netsurf, then Bajaj Finance

Brand and growth at national scale in India. Positioning as a discipline: distinctiveness is built through consistency, not isolated cleverness.

2021 → 2022 rePurpose Global

Global growth leadership at a mission-driven startup. Built and scaled the demand, automation, and attribution foundation.

2022 → 2025 Revinate

Led demand generation and lifecycle growth across inbound, nurture, and expansion, connecting experimentation and automation to measurable pipeline outcomes.

2025 → now Tigera

Global demand generation, North American SDR leadership, pipeline strategy, forecasting, and the operating model connecting marketing activity to revenue.

2026 → now Teamless

A passion project translating the campaign-memory thesis into a SaaS product.

The passion project that got serious getteamless.com

Every campaign produced results. Almost none left usable memory.

Teamless began as my attempt to fix that: a way to preserve the context, decisions, and learning that disappear between campaigns. Strategy in documents, execution in platforms, attribution in the CRM, lessons in someone's head. Every system captured its stage of the campaign. No system captured what the campaign taught.

V1 is built and in active use. I am now opening it to private-beta users while actively developing V2 with the intention of turning it into a SaaS product for demand generation teams.

Status: V1 built and in active use · opening a private beta · V2 actively in development

Visit getteamless.com ↗

Field notes SD · 2026

Essay · 01 GTM amnesia: the tax nobody budgets for
why organizational learning is a revenue problem, not a documentation problem

More field notes as they're written, not before.

Bring me the problem, not the brief.

If your pipeline is unpredictable, your funnel leaks in the same place every quarter, or your team keeps paying for the same lesson, I want to hear about it. The same goes for conversations about leadership roles, Teamless, or the systems behind this site.

Start a conversation LinkedIn ↗

shruti.dyundi@gmail.com · thoughtful problems welcome