Night School
Ask the Experts: The Tools We'd Actually Use Right Now
About this lecture
Everyone's shipping a new tool. Almost none of them are worth your time. Three Gauntlet AI practitioners tell you which ones actually are — then take your questions live.
New model, new agent framework, new “must-have” tool — every week. Cursor, Claude Code, and Codex are already colliding into one stack. MCP just shipped a breaking spec rewrite. Half the coding-tool debates online are vibe-coding backlash dressed up as best practice. Most of what's trending won't matter in a month, and figuring out which half is worth adopting eats time you don't have. On August 19, three Gauntlet AI practitioners — one in design, one in product, one in AI engineering — skip the theory. Each opens with the three tools they'd actually reach for right now, fast and opinionated. Then the floor is yours. Bring the tool you can't decide on, the one you dropped and wonder if you were wrong about, the workflow question you've been sitting on. No slides for the sake of slides. No vendor pitch. Just direct answers from people building with this stuff every day.
“Every week there's a new tool everyone's arguing about. This is the session where we just tell you what we'd actually use.” — Ash Tilawat, CTO, Gauntlet AI
In this lecture
- A working shortlist — three tools apiece from design, product, and AI engineering — that three practitioners actually reach for this week, not a "best of 2026" roundup
- A direct, on-record answer to the specific tool you've been unsure about, from someone who uses tools like it daily — ask it yourself in the live Q&A
- The actual questions these three ask before adopting anything, so you have a gut check ready the next time a tool lands in your Slack
- Three different disciplinary lenses on the same call — what a designer, a product lead, and an AI engineer each weigh when deciding hype versus worth-switching-to
- What each speaker downgraded or dropped this month, not just what's new — the unglamorous half of the picture most tool roundups skip
- A session built around your questions, not a deck — leave with a shortlist you can act on tomorrow, not slides you'll never reopen
About the speakers
Ash Tilawat — CTO of Gauntlet AI and architect of its curriculum, promoted from Head of Product in October 2025. He's trained 1,200+ engineers and open-sourced a 5-stage production eval framework (golden sets → labeled scenarios → replay harness → rubrics → experimental configs). Ash has spent years turning AI from a prototype toy into something teams actually ship — which is exactly the judgment call this session is asking him to make live, tool by tool.
Derek Peters — Professor at Gauntlet AI who came up through telecom and electrical engineering before moving into network engineering, Python, and full-stack development, then product — taking work from ideation through implementation and go-to-market. At Gauntlet AI he's inside instruction, grading, and product decisions daily, evaluating tools from both the builder's side and the "does this actually work for a team" side.
Claudia — Former ICU nurse turned software engineer with four years of building, breaking, and learning. Doodling on Procreate launched her into a design contest that won her a trip to Bogotá, Colombia, which led her to quit nursing and go all in on tech. She doubled down on her design interest, worked at several startups, and then launched her own fitness brand. With the help of AI she began building as much as she was creating, which led her to Gauntlet. Today she builds products for teachers at Solution Tree. Outside of work, she's learning to surf or exploring new design tools and building whatever sparks her curiosity.