thoughtbot

Product design and development consultancy

thoughtbot helps teams research, design and develop digital products. For Knect, it combined product discovery and mobile UX with experiments in AI-assisted recommendations.

Project minimum
$10,000+Clutch minimum ·
Hourly rate
$150 - $199 / hrClutch ·
Company size
50 - 249 peopleCompany-wide · Clutch
Listed location
New York, NY +1 locationClutch company listing

Start with the work

Published AI projects

1 reviewed project · Source-linked

Knect MVP user-flow diagram with connected screens and decision points, published in thoughtbot’s case study.View image
User-flow diagram · thoughtbot / Knect

Published project 01

Knect

AI-assisted social connection mobile app

Proof of concept

Agency contribution

Discovery research, UX/UI, a React Native proof of concept and feasibility experiments using custom GPTs and recommendation heuristics.

Delivery scope & limits

Proof of concept; not evidence of a production deployment.

Is this the right fit?

Services & project fit

Consider thoughtbot when the first task is deciding whether an AI product should be built and what can be tested cheaply. Knect combines customer research with technical spikes and a mobile PoC. Agree decision criteria for further investment and a separate scope for any production release.

  • Designs AI products
  • Builds AI systems

Services supported by reviewed work

  • Product discovery
  • Mobile UX
  • AI feasibility prototypes
Reviewed scope & sources

Discovery research, UX/UI, a React Native proof of concept and feasibility experiments using custom GPTs and recommendation heuristics.

Supporting sources: thoughtbot.com ↗

Full design scope & engagement notes

AI product design assessment

Early product discovery when both demand for an AI idea and technical feasibility need testing.

Design scope in the reviewed work

  • Interviews, assumptions mapping and user tests
  • Iterative prototypes and mobile UX
  • React Native PoC and feasibility spikes

Engineering responsibility

Knect includes application architecture, API experiments and GPT/SQL-heuristic exploration alongside design.

What the reviewed sources do not establish

Explicit proof of concept. No production rollout or independently validated recommendation quality is established.

Knect ↗

Engagement approach

Scoped project enquiry; discovery and build services. Numeric minimum and rates not stated on the reviewed enquiry page; request a quote.

Engagement source ↗

A question for your brief

What evidence would make us stop, change direction or fund a production build after discovery?

Fit and suggested questions are editorial guidance. Confirm the current service offer, delivery team and implementation responsibilities in the proposal.

Plan the conversation

Pricing & company details

Commercial terms

Clutch minimum project
$10,000+ minimum project · Clutch

Clutch-listed minimum for the company’s services, not an AI-specific quote. Confirm scope, currency and current terms.

Pricing source · Checked

Hourly rate
$150 - $199 / hr

Clutch-listed facts ↗ · Checked

An hourly band is not a scoped project quote.

Company & delivery

  • Company size50 - 249 people
  • LocationNew York, NY +1 location

Clutch-listed facts ↗Checked

Company-wide ranges and scale statements do not establish the size of your assigned delivery team.

View all 2 Clutch-listed locations

New York, NY · London, England

Official location / delivery coverage

Working model: Remote company; no physical offices. Official source ↗

How to interpret these listings

Clutch lists New York and London. The company’s own source describes remote working; directory locations should not be read as confirmation of staffed physical offices.

Availability

Confirm with the agency

Know what the record supports

Evidence & sources

How we research →
Detailed provider account

Evidence assessed

The provider’s AI-assisted social connection mobile app account describes specific work: Discovery research, UX/UI, a React Native proof of concept and feasibility experiments using custom GPTs and recommendation heuristics. The project contribution and delivery claims remain provider-reported; consult the stage and limitations below before shortlisting.

Assessment sources: Knect

Important limits

  • Knect is a proof of concept. Production scale, model ownership and independently validated recommendation accuracy are not established.

Profile sources

  • KnectProject accountthoughtbot.com · Checked

    agency published case

    Discovery research, UX/UI, a React Native proof of concept and feasibility experiments using custom GPTs and recommendation heuristics.

    Open original source ↗
  • Machine learning artificial intelligence aiCompany sourcethoughtbot.com · Checked

    supplementary primary source

    Company service or identity context.

    Open original source ↗
  • Start a thoughtbot projectCompany sourcethoughtbot · Checked

    agency service page

    Invites a scoped project enquiry.

    No numeric project minimum or rate on this reviewed page.

    Open original source ↗
  • thoughtbot handbook: remote companyCompany sourcethoughtbot · Checked

    agency company page

    Describes the company as remote with no physical offices.

    Open original source ↗
  • thoughtbot — company directory profileCompany sourceClutch · Checked

    directory profile

    Clutch-listed hourly rate: $150 - $199 / hr. Clutch-listed employee range: 50 - 249. Clutch-listed locations: New York, NY; London, England. Minimum project size checked 2026-11-14: $10,000+. This is a Clutch-listed starting threshold, not an AI-specific quote.

    Clutch lists New York and London. The company’s own source describes remote working; directory locations should not be read as confirmation of staffed physical offices.

    Only the new minimum-project field was refreshed in this pass. Other commercial fields retain their original snapshot date.

    Open original source ↗

Evidence labels describe the reviewed documentation, not a quality rating or endorsement. Company-reported delivery and results are not an independent audit. How evidence labels work.

Before you shortlist

Questions to ask thoughtbot

  1. What evidence would make us stop, change direction or fund a production build after discovery?
  2. Which parts of the experience did your team design?
  3. Can you share a current reference, confirm availability and quote against our actual scope?

Suggest a correction