Respell - No-code AI workflow automation for professionals

How It All Began

Every professional knows that feeling—the endless hours spent drafting documents nobody wants to read, reviewing memos that could write themselves, updating spreadsheets until your eyes glaze over, or writing code that's more tedious than transformative. These tasks are necessary, but they're not why you got into your field. The work that truly matters—building relationships, brainstorming new features, planning your company's future—keeps getting pushed to the back burner.

That's exactly the problem we set out to solve.

In late 2022, during the holiday break, our founder Matt Rastovac started working on something that seemed almost impossible at the time: a way to chain together multiple LLM prompts without writing a single line of code. Back then, having a computer think, search the web, and update a spreadsheet was frontier technology. The possibilities felt endless, but the tools didn't exist for most professionals to take advantage of them.

In January 2023, we officially launched Respell—an AI-first automation suite designed from day one to help professionals reclaim their time for what actually matters. Our mission was simple: let people focus on value-creating work while AI handles the rest.

The Journey That Followed

What started as a bold experiment quickly gained traction. By May 2023, we launched our Research Agent—one of the first AI agents in the industry that could autonomously conduct research tasks. This wasn't just another chatbot; it was a glimpse into the future of how AI could work for professionals.

Over the following months, we watched our platform handle millions of workflows for companies of all sizes and industries. We saw teams transform how they worked—automating document creation, streamlining memo reviews, eliminating manual data updates, and accelerating code development. The feedback was consistent: professionals were finally able to focus on work that actually mattered to them.

On January 17, 2024, our journey took an exciting turn. The entire Respell team joined Salesforce Agentforce, continuing our mission to build AI agents for every company. Our standalone product officially closed on March 1, 2024.

---

Core Features That Made It Possible

Building Workflows Without Code

The biggest barrier to AI adoption wasn't technology—it was accessibility. Most professionals couldn't hire developers to build custom AI solutions, and learning to code themselves wasn't practical. We solved this with our visual workflow builder.

Through an intuitive drag-and-drop canvas, users could connect multiple AI tasks into complete automated workflows. Need to research a topic, summarize findings, and create a presentation? Chain those steps together in minutes, not hours. The platform handled the complexity behind the scenes so users could focus on the outcome.

Research Agent: Our First Leap Into AI Agents

In May 2023, we launched Research Agent—marking our first step into the AI agent space. While others were still building simple chatbots, our users could delegate entire research tasks to AI. The agent would search, analyze, compile, and deliver actionable insights without constant supervision.

This wasn't just a feature; it was a statement. We believed the future of AI wasn't about answering questions one at a time—it was about giving AI autonomy to complete meaningful work.

Prompt Chaining That Actually Worked

Our prompt chaining technology allowed users to break down complex tasks into manageable steps, each handled by a specialized AI model. The output of one prompt became the input for the next, creating a seamless flow of intelligent processing.

Whether it was drafting a document, reviewing content for compliance, or generating code, users could orchestrate multiple AI capabilities without writing a single script. The system managed the logic, data flow, and error handling automatically.

Automating the Everyday

From day one, our focus was practical impact. Respell automated the tasks that consumed professional time but rarely moved the needle:

• Document creation: Generate reports, proposals, and communications in a fraction of the time
• Memo review: AI-assisted review that highlighted key points and flagged concerns
• Data updates: Automated spreadsheet and database updates that eliminated manual entry
• Code assistance: AI that handled repetitive coding patterns so developers could focus on architecture and innovation

---

Who Was Using Respell

Respell attracted professionals across industries who shared one common trait: they wanted to work smarter, not harder. Here's how different teams put our platform to work.

Marketing Teams Automating Content Creation

Marketing professionals used Respell to generate campaign copy, social media posts, and marketing materials at scale. Instead of spending hours drafting each piece, they created workflows that would take a brief and produce multiple polished outputs. The time savings let them focus on strategy and creative direction—the human elements that truly impact results.

Operations Teams Eliminating Data Drudgery

Operations managers dealt with constant data updates—spreadsheets that needed refreshing, reports that required compilation, status updates that consumed hours each week. Respell workflows automated these processes entirely. One operations lead told us they eliminated an entire day of manual work each week, redirecting that time toward process improvement initiatives.

Developers Accelerating Delivery

Even skilled developers have to write boilerplate code, refactor repetitive patterns, and handle mundane programming tasks. Respell's AI-assisted workflows handled these time-drains automatically. Developers reported they could deliver projects faster while focusing their energy on solving novel problems—the work they found most rewarding.

Researchers and Analysts Streamlining Discovery

With our Research Agent, analysts could delegate hours of information gathering and preliminary analysis. The AI would scan multiple sources, extract relevant data, and compile findings into actionable summaries. Research that previously took days could be completed in hours, giving professionals more time for insight and strategy.

Business Leaders Reclaiming Strategic Time

Perhaps most importantly, executives and managers used Respell to delegate the administrative burden that typically consumed their mornings. Drafting communications, reviewing routine documents, preparing status updates—these tasks now happened automatically, freeing leaders to focus on relationships, strategy, and innovation.

---

Technology Behind the Innovation

From Prompt Chaining to AI-Native Applications

When we started in late 2022, the AI landscape looked very different. Most "AI products" were simple prompt-response interfaces—useful but limited. We saw an opportunity to chain prompts together, creating workflows that could handle multi-step processes.

As the industry evolved, so did we. We watched the terminology shift from "prompt chaining" to "GPT wrappers" to "AI-native applications"—and we evolved with it. Our platform wasn't just wrapping GPT; it was building truly native automation that leveraged AI's strengths while hiding its complexities from users.

The Evolution from Chatbots to Agents

Our product journey mirrored the broader industry trajectory. We started with conversation-based interfaces, but quickly moved toward copilots that could assist with specific tasks. With Research Agent, we took our biggest leap into agent territory—AI that could work autonomously toward a goal without step-by-step human instruction.

This evolution wasn't just philosophical; it was practical. Each stage represented real improvements in what our users could accomplish. The millions of workflows our platform processed proved that professionals wanted AI that could do work, not just answer questions.

Technical Foundation

At its core, Respell was built on three technical pillars:

Visual Workflow Orchestration: Our canvas-based interface let users design complex processes without code. Under the hood, the system managed prompt sequencing, data transformation, error handling, and state management—everything needed for reliable automation.

Multi-Model Coordination: The platform could coordinate multiple AI models, each optimized for different tasks. A research workflow might use one model for web search, another for summarization, and a third for analysis. Users didn't need to understand the complexity; they just designed the flow.

Enterprise-Ready Reliability: We built from the ground up with reliability in mind. Workflows could handle real business processes, with proper error handling, logging, and integration capabilities.

---

Frequently Asked Questions

---

The Story Continues

A Year That Changed Everything

When we started building Respell in late 2022, AI automation was a distant concept for most businesses. The technology existed, but the tools to harness it didn't. Professionals were stuck doing work that machines could handle, leaving less time for the human work that truly mattered.

Our founder Matt believed there had to be a better way. During that holiday break, he built the first version of what would become Respell—a simple but powerful idea: what if anyone could connect AI capabilities together without needing to code?

By January 2023, we launched to the world. The response exceeded our expectations. Companies large and small, across every industry, saw the potential. They didn't want more AI tools to learn—they wanted AI that would simply make their work easier. Respell delivered exactly that.

Research Agent and the Future We Saw Coming

In May 2023, we launched Research Agent. Even in the rapidly evolving AI landscape, we knew this was different. We weren't just answering questions anymore; we were giving AI autonomy to complete meaningful work. Looking back, this was our first step toward what the industry now calls "AI agents"—and we were already there.

The months that followed validated our vision. Millions of workflows ran through our platform. Teams reported saving hours each week. The feedback was clear: professionals didn't want AI to replace their judgment; they wanted AI to handle the work that didn't require it.

Joining Forces to Scale Impact

When Salesforce approached us, we faced a meaningful decision. As an independent company, we were making progress but reaching only a fraction of the businesses that could benefit from AI automation. Joining Salesforce Agentforce meant our technology and team could impact every company, everywhere.

The choice was clear. On January 17, 2024, we announced our joining. Our standalone product closed on March 1, 2024, but our mission continues forward at Salesforce.

To Our Community

To the teams who trusted us with their workflows, who took a chance on an early-stage product, who shared feedback that made us better—we thank you. Your adoption proved that professionals everywhere were ready for AI that worked for them, not just around them.

The AI automation journey has just begun. What we started with Respell—making AI accessible, practical, and valuable for everyday work—continues through Salesforce Agentforce. The future we're building together will bring these capabilities to businesses of every size, in every industry, around the world.