Leaning into Agentic: How HoneySales Pivoted to an AI-Driven Future
Michael Cooper, Markus Kaiser and Ermin Dzinic talk with Denis Zatsepin, CTO of HoneySales, about how the company went from a manual sales service to SaaS and then to a multi-agent product on AWS. The agents build ideal customer profiles, a job that a sales consultant would otherwise do by hand, because customers with a wrong ICP got little from the rest of the pipeline. Denis built the first prototype with the Strands Agents SDK as one container on Fargate behind a load balancer, and had a working application in two to three weeks. An orchestrator agent proved slow and opaque to users, so the team moved to direct chats with specialist agents that sync through a shared workflow tool. Founders and engineering leaders get a concrete account of what a pivot to agents changes in architecture and user experience.
Special Guests
Denis Zatsepin
Founder and CEO of HoneySales, leading the company's pivot to an AI-driven, agentic sales platform.
Timestamps
00:00Welcome and the HoneySales pivot to agentic04:30Who is Denis Zatsepin and what does HoneySales do?06:30How HoneySales moved from a sales service to SaaS11:30How did customers react when the service became SaaS?14:30Why a wrong ICP breaks the whole sales pipeline16:30Why build an agent to replace the sales consultant?19:30Markus on agents moving the bottleneck away from users27:30How did HoneySales start building agents on AWS?30:30Strands Agents prototype on Fargate: start simple35:30From local prototype to multi-agent in two to three weeks39:30Orchestrator agent versus direct chat with shared workflow44:30What the panel learned: Strands SOP and Nova Sonic voice49:00Context engineering and cutting token use four to five times52:30Why an LLM is not a database: one dynamic agent55:30Agent as a tool and Nano Banana Pro for ThanksgivingAbout This Episode
- What is "Leaning into Agentic: How HoneySales Pivoted to an AI-Driven Future" about?
- Michael Cooper, Markus Kaiser and Ermin Dzinic talk with Denis Zatsepin, CTO of HoneySales, about how the company went from a manual sales service to SaaS and then to a multi-agent product on AWS. The agents build ideal customer profiles, a job that a sales consultant would otherwise do by hand, because customers with a wrong ICP got little from the rest of the pipeline. Denis built the first prototype with the Strands Agents SDK as one container on Fargate behind a load balancer, and had a working application in two to three weeks. An orchestrator agent proved slow and opaque to users, so the team moved to direct chats with specialist agents that sync through a shared workflow tool. Founders and engineering leaders get a concrete account of what a pivot to agents changes in architecture and user experience.
- What topics does this episode cover?
- This episode is filed under AI & Agents, SaaS Strategy, Founder Stories on Agentic SaaS Talks.
- How long is this episode, and when was it published?
- "Leaning into Agentic: How HoneySales Pivoted to an AI-Driven Future" runs 58 min and was published on November 28, 2025.
- Who is featured in this episode?
- This episode features Denis Zatsepin as a guest, in conversation with the Agentic SaaS Talks hosts.
- What are the key moments in this episode?
- Key moments include: Who is Denis Zatsepin and what does HoneySales do?; How HoneySales moved from a sales service to SaaS; How did customers react when the service became SaaS?; Why a wrong ICP breaks the whole sales pipeline; Why build an agent to replace the sales consultant?; Markus on agents moving the bottleneck away from users; How did HoneySales start building agents on AWS?; Strands Agents prototype on Fargate: start simple. The timestamps on this page jump to each one on YouTube.
Related Episodes
Episode 39: Enterprise Adoption of AI: When AI Starts Writing to the System of Record
Guest host Michael Cooper, founder of AGLedger.ai, joins Bill Tarr and Kamal Gupta to argue that the line that matters for enterprise AI is action, not adoption. He walks the adoption curve from chat and coding copilots through workflow automation, supervised action and bounded autonomy to probabilistic decisioning. The group covers agent identity, why an agent that resets its memory should be treated as a new employee on every run, and why results drift between a Monday and a Saturday. Cooper says enterprises already have four of the five pieces they need and are missing a durable, tamper-evident record of intent, approval and action, and why a log an agent reconstructs afterward is not evidence. Anyone building or buying agents that write to systems of record gets a concrete case for recording intent first.
Episode 38: The Last Mile of AI-SDLC: When Agents Build What Customers Want
Productmind co-founders Steen Brahe and Alessio Nobile take on the last mile of the AI software development lifecycle: getting from a real customer conversation to shipped code without losing the original intent. They demo how a customer call becomes a feature request, then a living specification, then agent-ready tasks running against an existing codebase, with full lineage preserved from feedback to release. The panel also digs into how AI is redrawing product and engineering roles, what governance of human-agent collaboration looks like in the enterprise, and how the founders cleared 20+ bugs with parallel agents while building Productmind with Productmind. For product managers it doubles as a working example of AI for product management: customer feedback turned into specs that coding agents can act on.
Episode 34: Agentic AI Governance in Production: Inside DataRobot's Agent Workforce Platform
Romi Datta, VP and Head of Product Management at DataRobot, explains why enterprise AI is moving past chatbots and copilots toward full agent workforces, and why 95% of agentic AI pilots still fail to deliver measurable ROI. He walks through production use cases from insurance claims processing to industrial safety at Chevron (with NVIDIA edge AI), then breaks down the governance, observability, secure context management, and agent memory an enterprise platform needs to run agents in cloud, hybrid, on-prem, and air-gapped environments.