# Two decisions. Pick the one that is stuck. > I take on two decisions: whether a stalled AI workflow should ship, and whether a business-critical Airtable system should stay or go. Both run through TechTide AI. Canonical: https://alexcinovoj.com/services I do not sell a menu of capabilities, and I do not sell sprints of general AI work. There are two decisions I take on, and both run through TechTide AI, where the scope and the pricing live. This page is the router. It tells you which engagement is yours and hands you to the page that spells out how it runs. ## The two decision engagements ### Systems Decision Audit An AI workflow blocked between a working demo and dependable operations. Returns one decision: Ship, fix, redesign, or stop. - A pilot that has been two weeks from shipping for two quarters - Agents with no evals, no traces, and no blast-radius limits - Model spend nobody can attribute to a workflow - Nobody willing to say out loud whether it is safe to trust Ninety minutes with the people who built it, then a written call: ship it narrow, fix a named list first, redesign the approach, or stop. The reasoning is on the page, so your team can argue with it. ### Airtable Rescue Decision Sprint A business-critical Airtable system that now decides what operations can do. Returns one decision: Stay, stabilize, hybrid, or exit. - Core operations living in bases nobody fully understands - Record, automation, and per-seat limits shaping process - No real access control on data that should be locked down - Reporting that needs a human to reconcile it every week A sprint through the real domain model, the limits you are actually hitting, and the cost of each option. It ends with stay, stabilize, hybrid, or exit, plus what the first move costs you either way. ## One step smaller, one step later. ## Common questions. ## Still not sure which decision is yours? Send me two sentences about what broke last week. If neither engagement fits, I will tell you that too. Technical proof lives on alexcinovoj.dev.