Requisor AI
Requisor AI is an applied AI company that builds its own agentic products, delivers custom multi-agent and voice systems for client teams, and trains organizations in the engineering disciplines that move pilots into production.
What is Requisor AI?
Requisor AI describes itself as an applied AI company, and on its services page as an AI services practice for brands and operators who want to ship something genuinely new rather than another chatbot. Its pitch rests on one statistic it repeats throughout: MIT research finding that 95% of enterprise AI pilots deliver no measurable return. The company's answer is that the model is rarely the problem, and that the context, the surrounding systems and the team's skills are. It organizes its offer into three modes: building its own products, building custom systems for clients, and teaching client teams to build.
The product line covers three tools. Requisor, the namesake product, is a context layer for product teams: calls, documents and prior decisions go in, structured context that both people and agents can build from comes out. It is live in beta. Concap AI is an agent for field sales that captures on-the-ground conversations through a wearable recorder, transcribes them with speaker separation, extracts the lead, pain points, objections and next steps, and pushes the result into a CRM. A third tool, an autonomous social content agent, is announced but not released.
The services arm delivers four lines — multi-agent systems, voice and conversational agents, streaming and generative brand experiences, and custom full-stack platforms — alongside analytics layers and brand and web work. Engagements follow a fixed sequence of discover, architect, build and operate, sold in three shapes: a two-week Strategy Sprint, a six-to-twelve-week full engagement, and an ongoing embedded partnership.
The training arm is anchored in a partnership with the Milwaukee School of Engineering, where the founder coordinates a graduate course, and extends into corporate workshops, multi-week cohorts and bespoke programs, organized around three disciplines the company calls prompt, context and harness engineering.
Three named client cases are published: Newaukee, where client diligence reportedly fell from six to eight hours to thirteen minutes; Toffi Talent, voice agents supporting workforce re-entry; and Peak Technologies, brand-adaptive streaming AI games. No pricing is published anywhere, and every commercial route ends at a fifteen-minute booked call.
What it does
- Design and ship multi-agent systems that run research, analysis, diligence and operations in parallel
- Build production voice and conversational agents for coaching, intake, screening and support
- Turn calls, documents and past decisions into structured, agent-consumable product context
- Capture field and trade-show conversations and push structured lead context into a CRM
- Create real-time, brand-adaptive streaming and generative experiences for activations and campaigns
- Build full-stack platforms, internal tools, customer portals and custom analytics layers with agents at the core
- Train engineering and non-technical teams in prompt, context and harness engineering
When to use Requisor AI / When not to
A quick filter to help you decide if Requisor AI is the right fit.
When to use Requisor AI
- Operators and brand teams that need a production-grade agentic system but have no in-house AI engineering bench
- Product managers who want meetings, documents and past decisions turned into structured context their tools can consume
- Field sales and trade-show teams that lose the detail of conversations between visits
- Engineering leaders looking to upskill a team in prompt, context and harness engineering through workshops or multi-week cohorts
- Universities and enterprise learning functions seeking an applied AI curriculum that ends in a working deliverable
When not to use Requisor AI
- Buyers who want to sign up and start immediately: every commercial route on the site ends at a booked call
- Teams that need published pricing before opening a commercial conversation
- Procurement and compliance functions requiring a privacy policy, terms of service or a data processing agreement up front, none of which the site publishes
- Developers looking for a public API or self-serve documentation
- Organizations that need a long, auditable track record: the domain dates from April 2025 and only one product is broadly available
How to use Requisor AI
A typical end-to-end flow, from setup to results.
- Identify which of the three tracks applies: an existing product, a custom build, or team training
- Book the fifteen-minute introductory call offered from every page, or write to the address matching your need
- Discover: a short call establishes the problem, the stakes and what success would look like
- Scope: a two-week Strategy Sprint audits your context, customer signal and stack, and returns an opportunity map, a build-ready roadmap and effort and ROI estimates
- Architect: the agents are defined, along with their tools, their handoffs and their guardrails
- Build: the system is delivered in two-week sprints over roughly six to twelve weeks, with your team in the loop throughout
- Enable: documentation, runbooks and knowledge transfer hand the system over to the internal team
- Operate: monitoring, tuning and expansion continue under an ongoing embedded partnership if you want one
- For training, book a session or ask for a custom program, and consult the graduate course listing in the partner university catalogue
- For the products, sign up directly on the Requisor beta or Concap web applications
Pros & Cons
Pros
- Three named client cases with published figures rather than anonymous references
- Unusually complete span: own products, custom delivery and training, with movement between them
- Verifiable academic anchoring, including a graduate course listed in a public university catalogue
- Explicit delivery cadence: two-week sprints, and six to twelve weeks to a production system
- Knowledge transfer built into the model, with documentation and runbooks at handover
- Founder-led engagement claimed from discovery through delivery
- The Concap product documents a consent-first approach, which is uncommon for audio capture
Cons
- No legal documentation whatsoever on requisor.io: no privacy policy, no terms, no legal notice, no DPA
- No published pricing, so no comparison is possible without going through a sales call
- No postal address and no complete legal entity name anywhere on the site
- Client-side site with no server rendering; the services and training routes answer HTTP 404 while still rendering in a browser
- Template placeholder text left in production inside the flagship case study, next to the figures it should support
- No social profiles, no robots.txt, and a training contact address on a different domain from the site
- Very young operation, with a domain registered in April 2025 and only one product broadly available
Pricing & Plans
Requisor AI publishes no pricing. There is no pricing page, no plan, no amount and no currency anywhere on the site, and neither a free plan nor a free trial is announced. Commercial entry is exclusively through a booked fifteen-minute call, the company stating that it sells outcomes rather than hours. The only pricing structure disclosed across the whole perimeter appears on the Concap product microsite, which describes a hardware lease combined with per-seat software and a volume discount for event teams, again without any figure. Prospective buyers should expect a quotation established per engagement.
- discovery
- two weeks
- delivers an opportunity map
- a build-ready roadmap and effort and ROI estimates. Price not published.
- build
- six to twelve weeks
- presented as the most common option
- delivers a production-grade system
- documentation and runbooks
- and knowledge transfer. Price not published.
- operate
- ongoing
- monthly tuning and expansion
- performance reporting and a new capability roadmap. Price not published.
- graduate course BUS 5900 at the Milwaukee School of Engineering. Price not published.
- corporate workshops
- half-day or full-day intensives on prompt
- context or harness engineering. Price not published.
- multi-week cohorts ending in a working production deliverable. Price not published.
- bespoke organization-wide programs built around the client's own stack
- datasets and compliance context. Price not published.
Data, GDPR & hosting
A consolidated view of how Requisor AI handles your data.
GDPR overview
requisor.io makes no claim of GDPR compliance. The single occurrence of GDPR on the entire site is a training outcome: the curriculum promises to teach teams to navigate GDPR, CCPA and the EU AI Act. No privacy policy, no legal notice, no Article 27 EU representative, no data protection officer and no legal or privacy contact address are published. The Concap product microsite does assert GDPR and CCPA compliance, along with two-party consent handling for US states and a SOC 2 Type II audit described as in progress, but it does so under a separate legal entity and its privacy, terms, DPA and security links are inactive placeholders. For a European buyer the practical position is that Requisor AI publishes no GDPR documentation at all.
Who owns the data?
Requisor AI publishes no terms of service and no privacy policy on requisor.io, so the question of who owns customer data is simply never addressed. The only ownership language anywhere in the perimeter sits on the Concap product microsite, which states that customer data stays with the customer, that audio is encrypted in transit and at rest, and that records can be deleted per contact on request. That page, however, carries a different legal entity in its footer, and its privacy, terms and DPA links are dead anchors. Ownership, licensing and any right the vendor may claim over material produced during an engagement therefore remain undefined and should be settled contractually before anything is shared.
Reuse rights
Nothing on requisor.io describes what the company may do with client or end-user data, because no terms of service exist to define it. There is no statement about whether customer data feeds model training, and no documented way to opt out. The Concap product microsite describes functional processing only: on-device transcription with speaker separation, automatic extraction of the lead, pain points, objections and next steps, and delivery of that structured record into the customer's CRM. Whether the end user may freely reuse the resulting transcripts, extractions and specifications, and under what licence, is not answered anywhere on the site and has to be negotiated directly.
Data retention & training
Hosting summary
Requisor AI declares no hosting jurisdiction. Because the site carries no privacy policy and no terms, there is no statement about where client or end-user data is stored, no named region and no named country, and nothing indicating whether European data could be kept in Europe. The only observable element is technical rather than contractual: the requisor.io domain resolves to an address operated by Google, reached through an anycast node geolocated in the United States. That resolution describes the marketing site itself and says nothing about where data processed in a client engagement, or inside the Concap product, would actually live. The Concap microsite mentions that audio is encrypted in transit and at rest and that retention is configurable, but names no storage location either. Everything else about the company points to a United States base, from the registrar to the university partnerships and the reference to US two-party consent states, without any of it amounting to a hosting commitment. Buyers with data residency requirements should treat hosting as entirely undefined and obtain it in writing.
Things to keep in mind
Risks and trade-offs to weigh before adopting Requisor AI.
- No published privacy policy, terms of service or data processing agreement: every data-handling guarantee has to be negotiated in the contract before anything is shared
- The exact legal entity is unknown, the footer showing only a trading name with no company form, registration number or address
- GDPR and CCPA compliance and a SOC 2 Type II audit are claimed on a product microsite operating under a different legal entity, whose legal links are dead
- An audit described as in progress is not a certification obtained, and treating it as one would be a mistake
- The headline 95% pilot-failure statistic is third-party research used as a sales argument, not a measurement of this company's own results
- Conversation-capture products record real people: consent tooling exists, but recording law varies by jurisdiction and the responsibility stays with the customer
- Placeholder template text left inside the flagship case study weakens the verifiability of the performance figures published beside it
Setup & Integrations
Technical difficulty
There is no self-service setup. Engagements are guided projects running through discovery, scoping, architecture, build and handover, with the client team involved throughout. First deliverable value arrives in about two weeks through a Strategy Sprint, and a production system in roughly six to twelve weeks, delivered in two-week sprints. Documentation, runbooks and knowledge transfer are included at handover, so the internal effort is mainly availability and subject-matter input rather than engineering. The products are lighter: the Requisor and Concap web applications are reached directly, though Concap additionally requires its dedicated wearable recorder.
Deployment
Integrations
Behind Requisor AI
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What exactly is Requisor AI?
Which products does the company publish?
How much does it cost?
Is there a free plan or a free trial?
How long does a project take?
Does Requisor AI offer an API?
Does the site publish a privacy policy or terms of service?
Is the company GDPR compliant?
Who teaches the training programs?
How do you get in touch?
Should you pick Requisor AI?
Requisor AI is a young, founder-led applied AI practice with an unusually complete offer: it ships its own agentic products, builds custom systems for clients, and teaches the engineering disciplines behind both. That span is genuinely rare, and parts of it can be checked from outside — a graduate course in a public university catalogue, three named client engagements, and published figures such as client diligence compressed from six to eight hours down to thirteen minutes.
The weakness is not the offer but the paperwork. requisor.io publishes no privacy policy, no terms of service, no legal notice and no data processing agreement. It gives no postal address, no complete legal entity name and no pricing. The site is a commercial showcase, not a supplier file. Buyers in regulated environments will have to obtain every contractual guarantee directly, and should note that the GDPR and SOC 2 language that does exist sits on a product microsite operating under a different legal entity, behind links that lead nowhere.
Execution details reinforce the impression of a business moving faster than its own website: the site is a client-side application whose services and training routes answer with a 404 status, and template placeholder text is still visible inside the flagship case study, immediately beside the numbers it is meant to support.
The reasonable conclusion is that Requisor AI is worth a conversation for organizations looking for an agentic build partner or applied AI training, particularly those that value a founder-led engagement and a real teaching practice. It is not yet a vendor that can be evaluated on documentation alone. Treat the first call as due diligence in both directions, and ask for the legal and pricing material the site does not provide.
- Choosing a selection results in a full page refresh.
- Opens in a new window.