Frequently Asked Questions
Common questions about Zach Varney's AI product management expertise, career, and approach to building AI products that actually work.
What products have I built?
I've led the development and launch of products across enterprise and personal projects. At Lytho, I spearheaded Insights, an AI tool that summarizes work items, drafts diagrams, and surfaces meaningful patterns, and Compliance, which enforces brand guidelines with AI-driven detection and actionable tasks. I also drove Slack and Microsoft Teams integrations and helped shape platform-level initiatives like Project Beacon, integrating DivvyHQ workflows into Lytho's broader ecosystem. At DivvyHQ, I guided major UI overhauls, bulk content management features, and built a data-driven culture with analytics tools like Pendo. Outside of work, I've built projects like Scanémon, a gamified card-scanning app, and AHVA, an experimental personal AI assistant with multi-layer memory.
What is my approach to AI product management?
I believe in being practical first, experimental second. My approach is to identify repeatable, high-value workflows where AI can remove friction, and design systems where humans remain in control but benefit from automation. I'm deliberate about scoping — I'd rather solve a problem deeply than spread thin with overpromises. I also invest heavily in infrastructure and modularity, so teams can iterate quickly without rebuilding foundations. Finally, I focus on cultural adoption, helping organizations not just ship AI, but trust it.
How do I evaluate AI model performance?
I look at a mix of technical and human-centered measures: Task success rate — can the model consistently complete the workflow? Accuracy vs. hallucination — if it doesn't know, it shouldn't guess. Latency and scalability — fast, reliable responses build user trust. Cost efficiency — every token matters, so optimization is part of the design. User trust and clarity — especially for non-technical users, the model should give confident, understandable answers or ask for clarification.
What companies have I worked for?
I'm currently a Technical Product Manager at Lytho, where I drive AI innovation, platform integrations, and strategic initiatives. Before that, I spent nearly eight years at DivvyHQ, where I grew from developer into Director of Product, overseeing the full product lifecycle for an enterprise content platform.
What technologies do I specialize in?
I work across the stack, but I specialize in: Backend: FastAPI, Python, PostgreSQL, Prisma, Firebase, Supabase. Frontend: React, Tailwind, PWA design. AI/ML: LLM orchestration, vector DBs, RAG pipelines, multimodal prompt engineering. Infra & DevOps: Railway, Docker, GitHub Actions, Cursor.ai IDE. Integrations: Slack, Microsoft Teams, iPaaS, Cloudinary, AWS (S3/CloudFront).
What's my leadership style?
I'm a hands-on product leader who thrives in tough situations. I've often been placed on 'trouble teams' to stabilize chaos and deliver results — and that's where I do my best work. My style is to challenge assumptions constructively, stay close to the technical details, and work shoulder-to-shoulder with engineers and designers. I balance being a builder, a connector, and a mentor.
How can they work with me?
The best way to work with me is collaboratively and iteratively. I believe in quick MVP slices, fast feedback loops, and co-creating solutions rather than tossing requirements over a wall. I'm comfortable setting strategy, but I don't hesitate to dive into details to unblock the team.
What makes me different?
I bring a dual background as both a developer and a product strategist, which means I can bridge vision and execution. I've proven I can shift organizations toward AI-first thinking, not just in theory but through shipped products that people use daily. Unlike many PMs, I also build and launch my own side projects, keeping me sharp and experimental. And through layoffs, pivots, and shifting priorities, I've built a reputation as the go-to PM who delivers when things get hard.
Still have questions?
If you didn't find the answer you're looking for, feel free to reach out directly.