Last Decade’s Product Is This Decade’s Feature: How Aurasell Is Rebuilding GTM From First Principles

Hey folks - Firas here.

This week’s PMF Playbook comes from my conversation on Inside the Silicon Mind with Jason Eubanks and Srinivas Bandi, founders of Aurasell AI. Both are former operators at Harness, where they helped scale the company from a single product into a multi-product, $100M+ revenue platform. Jason ran go-to-market. Srinivas ran product and engineering.

What made this episode unusually powerful wasn’t just the ambition - though replacing what looks like 15 legacy vendors in one platform is no small claim. It was the clarity of first principles, and how deeply product architecture, PMF, and execution velocity were intertwined.

This wasn’t a story about “adding AI to sales.”
It was a story about raising the floor of what a product must be in the AI era.

Let me walk you through what stood out.

Srinivas said something early in the conversation that framed everything:

Last decade’s platform is this decade’s product.
Last decade’s product is this decade’s feature.

That’s not rhetoric. It’s a warning.

Cloud, open source, and developer tooling already compressed the cost and time of building software. AI multiplied that compression. If you know how to design the right primitives, entire categories collapse into features almost by accident.

The implication for PMF is uncomfortable but real: if you’re building what would have been a great product five or ten years ago, you’re already late.

Aurasell exists because Jason and Srinivas believe the GTM stack is still priced and architected like it’s 2015, while buyer expectations are shifting to something fundamentally different.

They didn’t start by asking, “What should we build?”

They started by asking:

●     Who are the actual personas?

●     What are their real workloads?

●     What activities consume their time?

●     What workflows sit underneath those activities?

They identified five personas: SDRs, sellers, frontline managers, executives, and operations - and mapped the work behind each one.

Only then did they ask where an AI-native approach could:

This matters because most “AI GTM” products stop at step one. They wait for the user to ask. Aurasell’s thesis is that real PMF in AI comes from acting before the user knows to ask.

That’s not a UX improvement. It’s a different operating model.

At some point, they realized something unintuitive: by solving workflows end-to-end for these personas, they were encroaching on and replacing what looked like 15 separate legacy tools.

That wasn’t the goal. It was the consequence.

Here’s the key PMF lesson: they didn’t chase displacement; displacement emerged from coherence.

To make that work, they rejected the idea of sprinkling AI on top of a 20-year-old CRM architecture. They were both blunt about this: if your objective is context-aware intelligence and agentic automation, legacy frameworks become liabilities.

Instead, they built:

●     a unified data store

●     a clean metadata layer for context

●     an AI-native workflow engine

●     an agent builder

●     a performant UI that doesn’t feel like enterprise drag

The cost of this choice was four months of visible “nothing.”
The payoff was compounding velocity.

Once the primitives existed, building what people perceive as “15 products” became fast, sometimes minutes and extensible even by customers.

PMF here didn’t come from speed alone. It came from front-loading architectural correctness so velocity could compound rather than decay.

One of the most important insights Jason shared wasn’t about automation. It was about distribution of performance.

In most B2B sales orgs:

●     sellers spend ~30% of their time selling

●     10-12 tools are used daily

●     91% of orgs miss target

●     the top 25% of reps produce ~80% of revenue

The obvious play is productivity: automate admin, reduce tool sprawl, save time. Aurasell does that.

But the deeper problem is inconsistent execution at scale.

Managers can coach five reps closely. They cannot coach fifty or five hundred with the same quality. Enablement arrives too late. Feedback is generic. Performance gaps are diagnosed after the quarter is over.

Aurasell’s bet is that PMF comes from encoding excellence into the system itself:

●     guiding reps away from low-probability work

●     embedding proven sales frameworks into workflows

●     surfacing gaps while deals are alive, not dead

●     showing what “good” looks like with real examples

●     closing the coaching loop automatically

This isn’t about replacing managers. It’s about giving every rep access to what only the best reps and best managers used to provide.

Jason made an important clarification that’s easy to miss:

You can’t automate will. You hire for that.

That line matters.

Aurasell doesn’t claim to turn bad hires into great ones. What it claims is that once you’ve hired capable, motivated people, everything around them can be systematized:

●     decision quality

●     prioritization

●     coaching

●     feedback

●     pattern recognition

PMF here is not about magic. It’s about removing friction and variance so human potential actually shows up in results.

Their partnership is unusually clean.

Five years together at Harness. Hypergrowth. One to fourteen products. Single-digit millions to $100M+. No blowups. No ego wars.

What stood out wasn’t just harmony - it was role clarity and mutual respect.

Jason deeply understood the GTM pain because he lived it daily. Srinivas deeply understood how to build platforms that could move fast without breaking. Each trusted the other’s depth enough not to fake it.

There’s a PMF lesson here that often goes unsaid: founder-market fit is necessary, but founder-founder fit determines how fast you can compound it.

Both founders were clear on one thing: PMF is not something you “arrive at,” especially now.

AI shifts expectations continuously. Users become more capable. Tolerance for friction collapses. What felt magical last year feels table stakes this year.

Their response is simple but disciplined:

●     obsess over users, not competitors

●     keep looking around corners

●     raise the internal bar for what “great” looks like

●     treat PMF as a relationship, not a milestone

In Jason’s words, it’s unacceptable that people use self-driving cars and consumer-grade tech at home, then go to work and operate out of a 25-year-old database.

That emotional gap is as real as the technical one.

If I compress this episode into one idea, it’s this:

In the AI era, PMF belongs to teams who rethink the base layer, not the feature set.

Aurasell didn’t win by adding intelligence. They won by rebuilding the system intelligence sits on -  starting from personas, workflows, and first principles, then letting scale and displacement emerge naturally.

That’s not the fastest way to demo.
But it may be the fastest way to build something inevitable.

Until next time,

Firas Sozan
Your Cloud, Data & AI Search & Venture Partner

Find me on Linkedin: https://www.linkedin.com/in/firassozan/
Personal website: https://firassozan.com/
Company website: https://www.harrisonclarke.com/
Venture capital fund: https://harrisonclarkeventures.com/
‘Inside the Silicon Mind’ podcast: https://insidethesiliconmind.com/