# NeoAI

> NeoAI captures how a company's best people get AI work done and applies it to every task, so work comes back right the first time at the lowest cost that meets the quality bar. It runs in the customer's own cloud, behind their existing AI gateway, with no change to how teams work. Pricing is gainshare: no measured savings, no fee.

Last updated: 2026-10-25. Every number below carries a label: Estimate, Illustrative, Simulated, Public list prices, Worked example, Commitment, or Pilot: to measure. NeoAI has no published customer results yet; savings are measured per customer in a pilot.

## The problem NeoAI solves

- AI quality depends on who writes the prompt, so the same task gets different results across teams.
- Tasks take several attempts, and staff hours go to fixing and re-asking.
- Frontier models and deep reasoning get used on simple steps, on every attempt.
- AI cost = people time + (attempts × cost per attempt). NeoAI reduces all three: recipes cut attempts and fixing time, and smart delegation cuts the cost of each attempt.

## How it works

### 1. Distilled recipes (right first time)
- NeoAI watches how tasks play out and distils the shortest path to an accepted output into a recipe: the steps, the context, the tools and the checks.
- The next similar task gets the recipe automatically, and the recipe is updated after every run.
- A recipe earns its place by passing the customer's quality bar in the fewest attempts on real tasks. Public and in-house skill libraries are only starting material.
- Recipes are stored as building blocks. A related task borrows the blocks that fit, so only the domain-specific part has to be learned. Worked example: moving from resume screening to performance review reuses 4 of 5 building blocks. Learning carries over across related tasks only, and NeoAI shows which blocks it reused and how confident it is.

### 2. Smart delegation (lowest cost that meets the bar)
- A simplification agent splits each task into sub-steps and scores how hard each one is.
- Each step gets its own model tier (small, mid or frontier), reasoning effort, context and work budget.
- If a cheaper model fails a confidence or quality check, the step escalates to a higher tier automatically. The user gets a result, never an error.
- Public list prices show a gap of about 20x between frontier and small models (Public list prices).

### 3. Safe to adopt
- Deploys in the customer's VPC, behind their existing AI gateway. One endpoint change, no workflow change.
- Prompts never train anything for other customers.
- Recipes are owned by the customer and exportable. Every recipe version is logged and auditable.
- Security pack: DPA available; SOC 2 / ISO 27001 status to be confirmed.

## How NeoAI differs from AI gateways and model routers

Gateways and routers pick one model per request. NeoAI sits behind them and tunes the model, reasoning effort, context and budget for each sub-step, using recipes learned from the customer's own usage that carry over between related tasks.

## Evidence so far

- Simulated: in a replay of three open-ended agent tasks (market-entry research brief, bug fix in an unfamiliar codebase, churn analysis memo), the NeoAI-optimised agent used 65–72% fewer tokens and cost 62–69% less. These were not live model calls and quality was not measured. Token counts via tiktoken o200k_base, 21 Oct 2026.
- Estimate: NeoAI's savings estimator ranges from 55–70% lower spend for enterprise Q&A, 60–80% for engineering, and 65–85% for agents. A pilot replaces these with measured numbers.
- Pilot: to measure: cost per accepted task, attempts per task, staff hours on rework, first-pass acceptance, and quality pass rate.

## How to buy

1. Free audit (about 2 weeks): map where AI spend and rework go, and size the savings.
2. Pilot on the customer's own evaluation set (2–4 weeks): recipes learned, delegation tuned, quality tested on the customer's prompts against their acceptance criteria.
3. Roll-out with monthly re-tuning as models and prices change.

Pricing: gainshare, a share ([X]%, to be set) of measured savings. No savings, no fee.

## Who it is for

- Heads of AI and VPs of Engineering at AI-native scale-ups and mid-market companies spending $20k–$500k a month on LLMs.
- VPs and Directors of AI Platform at enterprises who need answers for the CEO, CTO, CIO and CFO.

## Links

- [NeoAI home](https://recipes.neoai.solutions/): the full product overview, savings estimator and audit request form
- Contact: contact@neoai.solutions