How I AI: Tech Recipes CEO Brian Chap Built a 'True Decision Engine'

Over nearly two decades in the advertising industry, Brian Chap has worked across every holding company, on the brand side at L’Oreal, and the sales side at Warner Bros. Discovery. Regardless of which side he was on, he noticed the same operations gap: disorganized, disjointed data. “The industry incentives shiny keys over digital fundamentals,” he explained. I understand standardizing taxonomy or tightening your operating model isn’t easy, but nothing great came from easy.” So in 2019 he started Tech Recipes, a strategic consultancy that creates custom operating models for brands and agencies. How has AI changed your business? To create these models, Tech Recipes consolidates workflows into a single engine by connecting systems, structuring data into a unified source of truth, and turning weeklong requests into instant deliverables like audits, RFIs, and execution templates. In 2024, Tech Recipes expanded its services into a product. RootIQ is a governed engine that turns fragmented cross-channel data into decision-ready answers so media agencies and in-house brand teams can align more quickly. “Think of it like ChatGPT for ad operations and analytics. Simply ask plain-English questions, get instant client-ready answers that drive revenue, decision velocity, and flag risk earlier,” Chap said. How was it built? RootIQ’s architecture logic and know-how were human built by applying Chap’s 20 years of expertise. The user interface is a modern web app on TypeScript and Next.js talking to a Python backend, and the chat interface is wired to answer based on live campaign data. Clear guardrails eliminate hallucinations, while standard sign-in and session protections keep access limited to authorized users. “We spent a year building RootIQ, knowing it would expand Tech Recipes into a true decision engine,” Chap said. “Today, it continuously improves itself by identifying issues, refining workflows, and recommending optimizations while delegating tasks across tools like Slack and Jira to move faster without sacrificing accuracy or control.” None of this works without the human element, however. Every change is manually reviewed to ensure quality control. “AI often behaves like people and takes shortcuts, which can introduce unwanted code, features, or functions, so it must be closely monitored,” Chap said. What does work look like? The Tech Recipes team has four full-time employees plus Chap, assisted by a network of freelancers. The staffers function as senior leaders from sales, customer experience, and development, working with their own teams of AI agents that facilitate the deliverables. Chap also built a custom operating model in a GitHub repository that streamlines onboarding, strategy, delivery, and reconciliation through structured markdown files. This setup lets him push new tools instantly to his team’s Cursor environments, enabling self-sufficiency while escalating only the most critical challenges. What is your monthly cost? Chap takes a modular approach to his internal tech stack that allows teams to prioritize only what they need, when they need it, instead of being locked into expensive all-in-one systems with unnecessary features. “I turn things on and off as I need them,” he said. He relies daily on tools like Claude Code to iterate and architect new features, then feeds them into Cursor, which connects across multiple LLMs through APIs. How do you use AI at home? Chap extended RootIQ into a personal chat advisor connected to his family’s medical records, birthdays, school forms, travel details, calendars, emails, and grocery lists. “Everyone in the house can open it on a phone and get a straight answer without me playing switchboard,” he said. Advice for getting started? "Treat it like a jigsaw puzzle. Find which LLM will be your first puzzle piece. Then ask it to help you create architecture for [insert your idea]. Learn what MCP is and how that piece fits your puzzle. Then try to break it as you continue iterating, exploring, and educating yourself." Brian Chap's AI Toolbox Claude ChatGPT Lovable Google Stitch Google Veo 3 Runway AI ElevenLabs

AI News

How I AI: Tech Recipes CEO Brian Chap Built a 'True Decision Engine'

Inside the strategic consultancy's RootIQ product, plus how Chap uses AI at home.

UNLOCK FULL ACCESS

Eva Kis

Brand Marketing

Amex Is Turning Harry Styles Fandom Into a Membership Perk

From Awareness to Action: Rethinking Mental Health Campaigns

The Great Brand Secret: Know Exactly Who You Are ft. Snap CMO Grace Kao

Why Scarlett Johansson's Low-Profile Beauty Brand Couldn't Last

Megan Fox Is the ED Treatment In New Ro Sparks Campaign

Join Gronk, Coach, The Home Depot, and More at ADWEEK House

Three NFL Wives Who Sacked the Stereotype and Ran Their Own Brand Plays

Ad Council PSA Reveals the Secret Struggles of New Motherhood

KFC Superfans Prove Their Devotion to Chicken in 30-Minute Game Show