---
title: "Self-Evolving Agents, Self-Improving ROI"
description: "Verified memory, governed locally, optimized for ROI."
author: "Saroop Bharwani"
authorTitle: "Founder & CEO, Senso"
publisher: "Senso"
owner: "Senso"
published: "2026-07-26"
updated: "2026-07-26"
license: "CC-BY-4.0"
canonical: "https://senso.ai/self-improving-roi"
also_at: ["senso.ai", "cited.md", "citeables.com"]
state: "Published"
topics:
  - verified learning loop
  - verified memory
  - self-evolving agents
  - governed memory
  - self-improving ROI
  - agent ROAS
  - verified context layer
citation: "Bharwani, S. (2026). Self-Evolving Agents, Self-Improving ROI. Senso. https://senso.ai/self-improving-roi"
---

# Self-Evolving Agents, Self-Improving ROI

![A luminous loop of light traced over an ocean of data points at sunrise — the verified learning loop](https://www.senso.ai/self-improving-roi/hero.png)

> Verified memory, governed locally, optimized for ROI.

By Saroop Bharwani · Founder & CEO, Senso · July 2026

## TL;DR

What 45 teams and 8 hours taught us about the economics of verified memory, and how we're funding what comes next.

## The missing piece

Six months ago, most agents completed tasks.

Today, they can remain inside a continuous loop: *set a goal → take action → observe the outcome → learn → act again.*

Models gave agents intelligence. Tools gave them the ability to act. Goals gave them direction. Loops gave them the ability to improve.

But something critical was still missing: **where does an agent's learning live? How do we know it should be trusted?**

That was the challenge behind the [Self-Evolving Agents Hackathon](https://luma.com/swarmhack) in San Francisco: *"build an autonomous agent that self-evolves in a continuous loop to achieve a goal — and takes real action. Publish, monitor, orchestrate, or transact, grounded in real sources."*

![Challenge slide: ship an autonomous agent that self evolves in a continuous loop to achieve a goal](https://www.senso.ai/self-improving-roi/challenge-slide.png)

*The challenge statement as the room saw it — with the ask that follows from it: publish the agent's output to cited.md, monetize it with agent payment rails.*

![Judging criteria slide: Autonomy, Idea, Technical Implementation, Tool Use, Presentation](https://www.senso.ai/self-improving-roi/judging-criteria.png)

*The rubric. Autonomy was the first criterion, and it asked the only question that separates a loop from a demo: does the agent improve without a human in the middle?*

In 8 hours, 45 teams shipped working agents. Of the 45 teams, five teams were awarded a Senso prize: SwarmAds, Ratchet, Immune, DeUna, and Greplica.

The strongest projects — in totally different domains — independently discovered the same architecture:

> Capture an outcome. Verify it. Store it. Reuse it. We call this the Verified Learning Loop.

## The Senso Loop

*How Senso helps developers unlock self-improving ROI*

![A raw signal streams in and closes into a continuous ring of light around a bright verified core — the Senso loop](https://www.senso.ai/self-improving-roi/senso-loop.png)

Every team used Senso the same three ways:

> **Self-improving ROI:** the marginal cost of a repeated outcome falls with cumulative verified usage.

What that buys you:

- **Costs fall as usage rises.** Every other agent system gets linearly more expensive the more you run it. This one gets cheaper.
- **You never pay twice for the same problem.** Solved once is solved for every future run — and for every other agent with access.
- **The savings are safe to take.** Nothing is kept until something proved it worked, so the cheap answer is also the checked one.

Here are some of the winning concepts built on top of the Senso platform. For each one, watch two things: **what it decided to keep**, and **what it stopped paying for.**

See it live at [senso.ai/verified-sources](https://senso.ai/verified-sources).

## Ratchet: the cost of intelligence falls with use

Ratchet, built by Aayush Dixit, was the clearest demonstration of self-improving ROI.

Coding agents forget everything between runs, so teams pay to fix the same bug again and again. Ratchet remembers a fix — but only after an independent test proves it worked. Same agent, same bugs, same models. The only variable was memory.

![AI calls per fix, memory on vs memory off](https://www.senso.ai/self-improving-roi/ratchet-memory-curve.png)

*With verified memory, AI calls per fix fall from 4.25 to 1 over five rounds. With memory off, they rise from 5 to 8.*

Then Ratchet pointed its memory at a different team's app — and fixed two of the first four bugs on contact.

![Cost per fix: $1.65 first encounter vs $0.03 every time after](https://www.senso.ai/self-improving-roi/ratchet-cost-per-fix.png)

*First encounter: four AI calls, $1.65, 3.2 seconds. Every time after: one call, $0.03, 0.8 seconds — re-tested by Replay.*

The agent didn't become more intelligent. **It stopped paying to think about something it already knew.**

**How they used Senso.** Every proven fix is saved to Senso with its full history — who found it, who verified it, how often it's been reused. Other agents can fetch that fix, trust it, and pay for it.

**The ROI story.** The same bug costs **$1.65 the first time and $0.03 every time after** — a 98% discount for remembering.

## DeUna: one answer makes the business permanently smarter

DeUna, built by Axel Cuevas, Anthony Lee, and Andrea Urbina, is a digital twin of a small-business owner who sells through WhatsApp — his voice, his slang, his negotiation style, learned from real conversations.

When something new comes up, the agent doesn't guess. A customer wanted to trade in a Samsung with a smashed screen; no policy existed, so the agent sent the owner three buttons: *Only this one / All the same / Reject.*

One tap: "All the same." The customer got an instant answer in the owner's voice — and every future customer gets it automatically. No re-ask. No guess. No drift.

The owner configured nothing. **He answered one message, and the business became permanently smarter.**

**How they used Senso.** Senso holds everything the agent is allowed to say — every price, policy, and promise. Each tap from the owner writes a new rule into Senso, permanently. Even the price floor lives there, so nobody can haggle the agent below it.

**The ROI story.** Sales close at 2 a.m. without the owner. **Every question he answers once, he never answers again.**

## Immune: every attack strengthens the defense

Immune, built by Subhan Poudel, applied self-evolution to agent security.

When an agent gets breached by a prompt injection, Immune reads the attack trace and synthesizes a guard rule — an "antibody." To reach production, the antibody must block the original attack, stop eight mutations of it, and leave twelve legitimate tasks untouched.

In testing, a sixth-generation attack compromised a frontier model 10 out of 10 times. Immune's antibody blocked it **10 out of 10 times.**

**How they used Senso.** Antibodies that survive the gauntlet are stored in Senso as versioned security records — a defense library every current and future agent inherits.

**The ROI story.** The ROI is the breach that never happened. **Every attack survived makes the next one cheaper to stop.**

## From ROI to ROAS

SwarmAds pointed the same loop at advertising: agents generate campaigns, test them in a simulated market, and promote only the winners into reusable memory. **The agent learns which actions create revenue and puts future effort there.**

**The ROI story.** Each generation of campaigns starts from the last generation's winners — **ROAS climbs with use instead of decaying.**

Greplica proved the underlying point. Same models. Same task. Different context. *Different outcomes.*

**How they used Senso.** The team ingested a repository into Senso and benchmarked four context strategies head-to-head — Senso-native included — with context as the only variable.

**The ROI story.** Better context, faster fixes, same models. **Context is an asset, not an input.**

## The next challenge: turn $2,500 into $25,000

A demo is the first step. We want funded businesses. Here's what's next for these amazing builders:

**Each winning team gets $2,500 in Senso platform credits to keep building with Senso and the Tokens& community — with one goal: demonstrate $25,000 in customer ROI or ROAS by San Francisco Tech Week in October.**

Ratchet: cumulative engineering savings. DeUna: recovered sales. Immune: breaches prevented. SwarmAds: measured ROAS lift. Not another demo — proof that self-evolving agents create compounding economic value.

## Not at the hackathon? The on-ramp is open.

The same ladder is open to any builder, starting today:

1. **Sign up —** onboard at [docs.senso.ai](https://docs.senso.ai) and get **$100 in Senso credits.**
2. **Prove —** turn it into **$1,000 of measured ROI or ROAS** — against a baseline, not a promise.
3. **Payments —** that result earns you an interview for a stream of **additional credits** to expand and grow your customer base with self-improving ROI.
4. **Payout —** hit $25,000 in customer value and we introduce you to **top-tier Silicon Valley investors** — with verified results instead of a pitch deck.

**Bonus:** get your brand mentioned on ChatGPT, Perplexity, and Google AIO using [cited.md](https://cited.md) to improve GEO metrics across frontier models. **An intelligent context graph is forming on the open internet, and these builders are among the first publishers.**

## Conclusion: the real meaning of self-evolution

The winning projects didn't retrain models. They built governed memory: successful actions became verified context that could be safely reused.

That distinction matters. Truth is **verified, attributable, versioned, and reversible** — otherwise "self-evolving" quickly becomes unpredictable and misaligned.

The next generation of agents won't win on intelligence alone. They'll win on what they can safely learn, remember, and reuse.

Since September 2024 we've run 20+ hackathons in San Francisco with tokens&, drawing more than 20,000 developers. We keep the challenge statement [deliberately consistent across events](https://x.com/saroopbharwani/status/2011928481761018000) — small changes in wording have outsized effects, and adding one word like "self-improving" is enough to move a whole room from static agents to feedback loops.

The wording barely moves. **The goal never has: get someone funded.** That's what $2,500 into $25,000 is — the same goal we've always had, finally written down as a number, with verified results standing in for the pitch deck.

> The model provides intelligence. The loop produces experience. Verified memory turns that experience into compounding ROI.

## Start your onboarding

Three doors into the loop. Pick the one that matches how you build.

**For agents — give your agent the CLI.** Hand the Senso CLI to Claude Code, Cursor, or Codex and it installs the skills, ingests your sources, and starts writing verified memory on its own. → [docs.senso.ai/docs/senso-for-agents](https://docs.senso.ai/docs/senso-for-agents)

**For developers — track a domain, own the answer.** Point Senso at a domain — [cited.md](https://cited.md), for example. That's the GEO workflow: watch how ChatGPT, Perplexity, and Google AI Overviews answer for your org name, close the gaps, re-measure. → [docs.senso.ai/sign-up](https://docs.senso.ai/sign-up)

**For your industry — claim an industry.** Every industry gets its own tracked prompt set and leaderboard. Sign up for yours, or request one if it isn't live yet. → [geo.senso.ai/industries](https://geo.senso.ai/industries)

Whichever door you pick, the loop is six commands. Your agent can run them for you:

```bash
npm install -g @senso-ai/cli
senso login
senso skills install --all
senso ingest <your-domain>
senso compile --verify   # the gate
senso generate run
```

---

**Saroop Bharwani** · Founder & CEO, Senso · saroop@senso.ai · [senso.ai](https://senso.ai) · [docs.senso.ai](https://docs.senso.ai)

*Senso is the verified context layer for self-evolving agents — where what your agents learn from humans becomes governed, reusable truth. If you're a developer, get started with $100 in credits at [docs.senso.ai](https://docs.senso.ai), or give our CLI to your agent: [docs.senso.ai/docs/senso-for-agents](https://docs.senso.ai/docs/senso-for-agents)*

Cite as: `Bharwani, S. (2026). Self-Evolving Agents, Self-Improving ROI. Senso. https://senso.ai/self-improving-roi`