# Assist — Build AI Teammates

> Assist allows you to build cost-optimized AI teams that connect to your internal tools and data and operate inside your business. Managed AI teammates, connectors, Sandcastles, permissions, and model choice for teams that need AI in production.

Assist is built by Assist Technologies LLC. Get started at [app.withassist.xyz/sign-up](https://app.withassist.xyz/sign-up).

## Table of Contents

- [What Is Assist](#what-is-assist)
- [The Problem Assist Solves](#the-problem-assist-solves)
- [Model Routing](#model-routing)
- [Use Cases](#use-cases)
- [Where AI Teammates Work](#where-ai-teammates-work)
- [Connectors](#connectors)
- [Governance and Permissions](#governance-and-permissions)
- [Cost Management](#cost-management)
- [Case Study: People's](#case-study-peoples)
- [Pricing](#pricing)
- [Get Started](#get-started)

---

## What Is Assist

Assist is a platform for building AI teammates: AI agents that join your team's real workflows, connect to your internal tools and data, and operate inside your business under shared permissions. An AI teammate participates in channels and threads alongside humans, uses company systems through governed connectors, and hands work back and forth with people — it is not a standalone chatbot.

Examples of AI teammates in action:

- **Customer follow-up** — an AI teammate preps a customer meeting by checking Google Calendar, HubSpot, and Zendesk, then revises an email draft based on feedback from the Head of Sales and Customer Success in a shared channel.
- **Personal teammate** — a personal agent drafts replies in the operator's voice using saved memory of tone, approval habits, and customer preferences.
- **Finance close** — an AI teammate compares monthly software subscription spend in QuickBooks and Google Sheets, flags renewals needing approval, and sends summaries via Slack.
- **Hiring ops** — an AI teammate ranks candidates from Lever against a scorecard and places interview holds in Google Calendar.

Assist plugs into Snowflake, NetSuite, Shopify, Gmail, Salesforce, Postgres, Slack, and any API.

Trusted by: Medalist DST, People's California, Creamistry, PK4 Media, Fort Ashford Funds.

## The Problem Assist Solves

AI started as single-player, but work is multiplayer. Teams are already building with prompts, skills, AI agents, and local workflows — but each person does it individually, on their own laptop, with their own credentials, and no shared record. Assist turns that scattered activity into a shared operating layer where the human team and agents use the same systems, follow the same permissions, and hand work back and forth safely.

Without Assist (single-player AI):

- Each person prompts, scripts, and wires AI differently
- Access follows individual credentials instead of shared policy
- Provisioning and de-provisioning depends on local setup
- Human work and AI work stay disconnected from team operations

With Assist (multiplayer AI):

- Human and AI teammates work from the same governed systems
- Role-based permissions apply before data is read or actions run
- Shared workflows can be provisioned, audited, and improved centrally
- Teams hand work back and forth safely

## Model Routing

Assist does not tie your business to one model provider. Each AI teammate runs on the best, most cost-effective model for its workflow: routine work routes to cost-optimized open models, and frontier models are used only where they matter.

Supported model families span three tiers:

- **Frontier:** Claude (Anthropic), Gemini (Google), Grok (xAI), GPT (OpenAI), Perplexity
- **Open weights:** Mistral, Gemma (Google), Qwen (Alibaba), Kimi (Moonshot AI), GLM (Z.ai), Cerebras
- **Your hardware:** Llama (Meta), DeepSeek — including on-premise and self-hosted deployment

Small, fast, in-house models handle simple high-volume work; capable, cheap open models handle everyday reasoning; the most capable frontier models handle complex, high-stakes tasks.

## Use Cases

Assist lets teams scale without adding headcount by turning support, renewals, collections, ecommerce, and manufacturing workflows into AI teammates:

- **Resolve customer issues faster** (Customer Ops) — resolve billing, order status, plan changes, and refunds; troubleshoot common issues end-to-end; escalate edge cases with full customer context.
- **Protect renewals before they slip** (ARR Protection) — watch usage, tickets, sentiment, and renewal dates; trigger save plays before accounts go dark; draft renewal plans and CSM nudges.
- **Get customers live faster** (Onboarding Ops) — turn signed contracts into live customers; provision accounts, configure settings, and migrate data; track training, integrations, and time-to-value.
- **Collect cash sooner** (Finance Ops) — track unpaid invoices and payment promises; send follow-ups and reconcile incoming payments; escalate disputes and high-risk accounts.
- **Recover missed revenue** (Found Money) — compare contracts, usage, entitlements, and invoices; find underbilling, missed true-ups, and expired discounts; surface recovered revenue with source evidence.
- **Win complex deals faster** (Procurement Ready) — qualify inbound deals and build custom proposals; fill RFPs and security questionnaires; prepare ROI models and procurement follow-ups.
- **Prevent production delays** (Manufacturing Ops) — catch shortages before a production line stops; coordinate ERP, WMS, maintenance, and supplier updates; escalate recovery plans with approved next actions.
- **Protect ecommerce margin** (Margin Protection) — watch SKU margin, promos, inventory, and fulfillment cost; catch discount leakage before campaigns launch; draft merchandising changes with source evidence.
- **Ship custom apps without SaaS sprawl** (Custom Apps) — build CRM, ops, finance, and support apps around your workflow; combine AI teammates, forms, dashboards, and approvals in one Sandcastle; govern every screen and action with Assist permissions.

## Where AI Teammates Work

Assist AI teammates meet your team where work already happens — there is no new app to adopt. Assist sits in the middle as a hub, routing each surface to the right AI teammate and model.

Supported surfaces:

- **Slack** — channels and approvals
- **Microsoft Teams** — team chat and escalation
- **Browser** — internal apps
- **ChatGPT** — assistant surface
- **Claude Code** — developer workflow
- **Codex** — code teammates

Behind the hub, named AI teammates (Research, Outreach, Pipeline, Support, Reporting, Scheduling) each run on a model chosen for the job.

## Connectors

The Infinity Connector connects your AI team to the APIs and EDI systems behind each workflow, then wraps every action in policy. Connectors are owned by the company at the platform layer — not wired up on individual laptops — and every connection is scoped, rate-limited, and revocable.

## Governance and Permissions

Every action the AI team takes is checked against role-based policy before it can read or write data. An AI teammate can work across company systems — ERP (NetSuite, Oracle), CRM (Salesforce), data warehouse (Snowflake), inventory (WMS), GL/finance (Sage Intacct), email, documents, partner APIs, and analytics — while restricted systems (for example, HR data in Workday) are blocked by policy.

- Policy is checked before every tool call
- Restricted systems return "access denied" rather than data
- An immutable audit trail records every action, including denials
- PII redaction and rate limits are enforced at the gateway

## Cost Management

Assist breaks spend into tokens, tool calls, LLM spend, and tool spend, so teams can see what a workflow costs, compare it to business impact, and optimize usage — calculating the ROI of every workflow.

- **Token spend** — total token costs broken down by AI teammate instead of hidden in a blended platform bill, so model costs become something operators can actually tune.
- **Tool spend** — connector and action costs tracked alongside model usage; every workflow shows model spend, tool spend, total, and share of the monthly bill.
- **Cost visibility** — find expensive runs and agents, and tighten workflows before usage drifts; see which AI teammates, models, and tool calls drive spend so teams know where to optimize.

## Case Study: People's

People's built real operational software on Assist without a software development background.

> "Assist has been the scaffolding that we needed, as people who did not have any software development background, to build exactly what we were looking for."
> — Bradley, Head of AI Projects, People's

Before: an expensive ERP that didn't fit the business, with holes filled by dozens of spreadsheets. After: core ERP AI teammates and modules built by subject-matter experts.

How it happened:

1. People's had exact operational workflows to build but no software development capacity in the team.
2. Assist shipped a governed AI teammate into the team's workflow so subject-matter experts could start building with it.
3. The team learned to operate and extend it while Assist handled roles, permissions, audit, and safe execution.

**Result: 20+ internal tools at 90% less cost** — highly customized operational assets built around their exact workflows, governed by the Assist platform.

## Pricing

Whether Assist builds your AI team for you or your human team builds its own, there is a plan. No seat fees — unlimited users on every plan.

- **Managed AI Team — $10,000/month** ("We build it for you"). Everything in Self Serve, plus three AI teammates built and managed for you by an embedded engineer who designs, ships, and tunes them around how your team works. Includes on-premise LLM or AWS Bedrock support, first AI teammate live in under two weeks, monthly review and optimization, and priority support.
- **Self Serve — $20/month.** Start building immediately with no commitments or minimums. Unlimited internal SaaS apps, workflows, AI teams, tools, and user seats; access to OpenAI, xAI, Anthropic, Perplexity, and Cerebras models; role-based access control; tool call isolation.
- **Enterprise — Custom.** Everything in Managed AI Team, plus Single Sign-On (SSO), Directory Sync (SCIM), BYOC or on-prem hosting, and audit logging.

## Get Started

Assist turns scattered AI work into governed production workflows, so teams can deploy AI teammates that use company systems safely, choose the right models, ship internal apps, and control costs.

- Sign up: https://app.withassist.xyz/sign-up
- Documentation: https://withassist.xyz/docs.md (machine-readable table of contents)
- LLM summary: https://withassist.xyz/llms.txt

connectors · permissions · model choice · Sandcastles · cost visibility

© 2026 Assist Technologies LLC