Runtime vs Instinct: Personal AI Agent vs Team Ops

Feature	Runtime	Instinct
What it is	Agent harness for payment and fintech teams	Personal AI assistant you text or call
Who it works for	Teams across the org	One person
Vendor dependency	Multi-cloud, multi-sandbox, multi-model	Hosted by Instinct, models not named
Where agents run	Your AWS, GCP, Azure, or self-hosted	Instinct's hosted service
Guardrails	Read-only start, approvals, RBAC, masking	User is responsible for actions
Training on your data	Your models, your cloud	On by default, opt-out available
Team collaboration	Shared agents, templates, and memory	Not offered
Phone calls to businesses	Voice channel for agents	Instinct Concierge, early access
Setup	Days, with a forward-deployed engineer	Text it, once invited

Instinct and Runtime at a glance

Instinct is a personal AI assistant from Spear Street Technology, Inc., founded by Noah Shinn, a former research scientist at Sierra. You text it or call it. It connects to your email, messages, calendar, and other apps, and uses a phone and a computer the way a person would: disputing a bill, sending gifts, ordering groceries, booking a doctor's appointment, planning a weekend away. In September 2026 it added Instinct Concierge, which places phone calls to businesses for you, in early access. It is invite-only, for adults, and has drawn very large funding rounds in 2026. It is for one person who wants their errands handled.

Runtime is the AI agent harness for payment and fintech teams. Payment ops, risk, compliance, finance, underwriting, onboarding, and support build agents from their SOPs. Each agent works on its own isolated computer in your cloud, reaches your ledger, processor, and bank portals through APIs, databases, MCP servers, and a browser, and stops for a person before anything moves money. It is for operations leaders who run many agents across the org.

Good products, different jobs. The honest point: a payments team should not run operations on a personal assistant.

How they differ

No single-vendor dependency

Instinct is a hosted service. Its privacy policy says it relies on third-party AI model providers but does not name them, and you cannot choose where it runs or which model it uses. That is fine for booking dinner.

For operations, it is a dependency you can't manage. Runtime is multi-cloud (agent computers in your AWS, GCP, or Azure account, or fully self-hosted with Helm), multi-sandbox (bring your own sandbox provider, with fallbacks), and multi-model (Anthropic, OpenAI, Google, and open-weight models served in your cloud). It routes across harnesses such as Claude Code, Codex, and OpenCode. If one provider has an outage, changes pricing, or deprecates a model, your agents keep running.

Guardrails you set, not terms you accept

Instinct's terms state that you are responsible for its inputs, actions, and outputs, "including all consequences, whether financial, contractual, legal, reputational, or otherwise." The terms also allow your materials to be used for model training by default, with an opt-out. Both are reasonable for a consumer product. Neither works for a company holding card data and a sponsor bank relationship.

On Runtime, your team sets the guardrails. Agents start read-only. Approvals sit in front of releasing a payout, applying a reserve, booking a ledger entry, or replying to a sponsor bank. Roles control who can build, run, and approve. Credentials are masked, and card numbers and SSNs are stripped from prompts and logs. Every run is recorded end to end, from trigger to each query, tool call, approval, cost, and result, and the record your sponsor bank asks for stays with you.

Built for one person vs built for teams

Instinct is organized around you: your apps, your location, your messages. There are no shared agents, roles, or spend controls for a team.

Payment work rarely belongs to one person. A single stuck payment touches four teams: support gets the ticket, payment ops traces it, finance sees a recon break, and risk or compliance may review it. Today each team investigates it separately in different tools. On Runtime it is one investigation with one record. Teams share agents, templates, and memory, work on the same runs in real time, and manage access and spend by role.

Fintech depth and an engineer to get live

Instinct is self-serve once you have an invite. Runtime pairs you with a forward-deployed AI engineer whose team built payment and fintech infrastructure at Hulu's payments team at Disney, Finix, Modern Treasury, and BlackRock's AI quant group. The FDE maps your processes, builds the first agents with your team, sets up guardrails, and trains admins. Once a process is solved, agents can turn it into a script that runs deterministically, in your repos.

Where Instinct is stronger

Personal errands. Bills, gifts, groceries, appointments, and travel for one person.
Phone calls to businesses. Instinct Concierge calls places that don't take online bookings.
No new interface. You text or call it like a person.
Zero setup. Once invited, there is no implementation project.

Pricing

Instinct does not publish pricing. Its terms say some features may require fees, charged in US dollars and non-refundable except where required by law.

Runtime has a Free tier ($0, one session), Teams from $99 per seat per month, and Enterprise with custom pricing and self-hosting. Most teams measure it against the work of analysts they were about to hire. Rain replaced $250k in vendor spend with Runtime.

Which should you choose

Choose Instinct if

You want a personal assistant for your own errands, bills, and bookings
Texting or calling an assistant fits how you work
The data involved is your own, not your company's customers'

Choose Runtime if

Agents need to work card data, PII, ledgers, or payouts
You can't depend on one cloud, sandbox, or model vendor
Your team, not a vendor's terms, should set the guardrails
Several teams touch the same transaction and should share one investigation
You want an engineer who has built payment infrastructure to get the first agents live

See Runtime on your busiest queue

Bring one SOP. A forward-deployed AI engineer builds the first agent with your team, inside your cloud.

Book a demo

Frequently asked questions

What is Instinct AI?

Instinct is a personal AI assistant from Spear Street Technology, Inc. that you reach by text message or phone call. It connects to your apps and devices and uses a phone and a computer to do tasks such as disputing a bill, ordering groceries, booking appointments, and planning trips. It launched as an invite-only beta in August 2026.

Can a fintech team use Instinct for payment operations?

Instinct is built for one person's life admin, not for team operations on card data, PII, ledgers, or payouts. Its terms make the user responsible for every action it takes, and it does not offer team roles, shared agents, or deployment in your own cloud. Runtime is built for that job.

Does Instinct train on my data?

Instinct's terms say it may use your materials to train AI models by default, with an opt-out in settings. Content stored in its Vault feature is excluded, and its privacy policy says Google Workspace data is not used for training. With Runtime, agents run in your cloud on the models you choose.

How much does Instinct cost?

Instinct does not publish pricing. Its terms say some features may require fees. Runtime has a free tier, Teams from $99 per seat per month, and custom Enterprise pricing with self-hosting.

What is the best Instinct alternative for business teams?

If you need agents that work for a team rather than one person, on regulated data, with approvals and an audit trail, Runtime is built for payment and fintech teams. For personal errands, Instinct, Meta Muse, and OpenAI Dots are the closer comparisons.