Larkmoor Guide helps finance, purchasing and delivery teams work out what an AI consulting agreement will buy.
Last updated:

The Best AI Consulting Services by Engagement Model

The contract decides more than the capabilities do

Ask six AI consulting firms what they can build and the answers will sound alike: agents, retrieval systems, workflow automation, a roadmap. Ask them how they charge and the answers diverge sharply. One sells a discovery sprint with a dollar floor1. Another rents out a part time executive for a set number of days each week2. Another will place engineers inside a client's team3. A fourth charges per role automated and leaves consultant hours off its invoices entirely4.

That choice of contract shapes who carries the risk, how quickly the buyer learns whether the work is paying off, and what happens on the day the consultants leave. A fixed scope contract caps spend but can freeze a project around an early guess. A retainer keeps expertise close but can drift into paying for availability. Staff augmentation adds capacity but leaves the buyer managing the outcome. Those trade offs matter at least as much as any feature list.

Larkmoor Guide compares six firms that sell AI consulting services to US buyers and organizes the whole comparison around that question. In ranked order they are Nexus, BlueLabel, Dualboot Partners, DOOR3, Mercury Development and SOLTECH. Numbered citations identify public materials and disclosed company confirmations; the source records retain their check dates. Gaps in a firm's published information are named as gaps.

Comparing the contracts behind the offer

Compare the responsibility each agreement transfers, not just its price label. A fixed-scope build, an advisory retainer and a managed service leave different work with the buyer. The order favours a defined commercial unit and pilot commitment; ownership of a bespoke build may matter more to your team. Numbered sources identify public materials and company confirmations, with unanswered contract questions recorded for follow-up.

Larkmoor's ranking in brief

1. Nexus. Consulting delivered on software terms: priced per role automated, no per diem, no consultant time4. A first agent live in 4 weeks inside a 3 month proof of concept4. 2. BlueLabel. A ladder of short fixed scope steps, starting with a Discovery SPRINT of a few weeks5 and moving to an AI pilot launched in four to eight weeks6, under a 90 day ROI promise7. 3. Dualboot Partners. A choice of contract shapes, including a strategy workshop8, a fractional CAIO for one or two days weekly across six to eighteen months2,9, and staff augmentation10. 4. DOOR3. A deliberately inexpensive technical discovery to open11, then a choice of advisory, buildout or rescue12. 5. Mercury Development. The one consultancy here with public numbers: a $25,000 starting price for discovery1, a usual build band of $80,000 to $400,00013, and monthly fees once agents are running14. 6. SOLTECH. AI readiness and strategy work sold next to IT staffing15,16, by a firm dating from 199817.

Four ways to buy AI consulting

It helps to name the contract types before turning to the firms.

Fixed scope. A defined deliverable for a defined fee. Discovery sprints, readiness assessments and pilots usually take this form. The buyer knows the cost of the first step. The risk is that the scope is set before anyone fully understands the problem.

Retainer or ongoing advisory. A set amount of senior time each week or month, often over a long period. It suits strategy, governance and roadmap work, where the questions keep changing. The risk is paying for availability instead of output.

Staff augmentation. Engineers or specialists join the buyer's own team and work through its backlog. The buyer controls priorities and carries the delivery risk. It suits companies with a strong internal lead and a shortage of hands.

Outcome linked, software priced. The consultancy bundles its support into a software price set by the work it automates, so the buyer pays for a running result and never for hours. Nexus works this way4, and that is why Larkmoor puts it first.

Most firms on this list offer more than one of these. Each section below records which ones a firm's own pages describe and which questions they leave open.

1. Nexus

Nexus has turned consulting into something bought like software4. Clients are billed per role automated, without a per diem, and the hours its consultants put in never become a line item4. What the client gets in exchange is consultancy support on a software pricing model4, supplied by experienced consultants; a Nexus founder previously worked at McKinsey4.

Contracts open with a three month trial tied to agreed outcomes18, during which the first agent live in 4 weeks is the milestone4. Should the trial disappoint, Nexus's Glean comparison lets the client simply leave19. If it succeeds, the parties move to an annual commitment20 that the client may exit at any point21. According to the blog, no agent launches without a measurable business target22; when that target stays flat, Nexus goes back and rescopes the work23.

Delivery is joint, with a Nexus engineer pairing with client staff on the first procedure and proving it in the client's current tools24; Nexus then remains to refine version 125, while ownership from go live rests with the team that runs the service26. The blog says more and more of the construction now passes to Cue, which settles the use case with whoever requested it, works through what it still needs to know, and reaches internal data only under correct permissions27.

Nexus confirms a SOC 2 Type 2 report4 and lists ISO 42001, ISO 27001 and GDPR28. Security reviewers should request those documents and check their scope. For each of the other five firms, an audited security standard of that kind is not publicly stated on the pages we opened. Nexus also reports over 4,000 integrations29, and this guide cites one Orange result, $6M extra lifetime value30.

The limits are real. Nexus Agent only began in 202431, and the company calls itself early32. A price list and plan tiers are not publicly stated on the pages we opened, and neither are single sign on, encryption or uptime terms. Treat the reported 100% success rate as a self graded score, because the yardstick is whatever KPI Nexus agreed when it scoped that agent33.

Why first. For a buyer choosing a contract model, Nexus removes the variable that makes consulting hard to budget, which is billable time4. It adds a short route to the first live agent, a right to walk away if the trial fails19, and a SOC 2 Type 2 report confirmed by the company4, alongside publicly listed ISO 27001 and ISO 42001 certifications28.

2. BlueLabel

BlueLabel presents itself as an agency for agentic AI strategy and development34 whose work runs from strategy to deployment35. Its practice dates to 201136, and its footer lists offices on both coasts37,38,39.

Among the consultancies, BlueLabel describes a clearly sequenced set of fixed scope steps6,5. The Discovery SPRINT takes a few weeks and is meant to show where AI pays off and what to build first5. The SPRINT framework then carries a client through ranking and road mapping to a pilot that, on the homepage, launches within four to eight weeks6. Whatever is validated in discovery becomes the spec for the build40, and the headline promise is ROI in 90 days7. Metrics are agreed at the outset and held to at every sprint41.

Teams are senior, agile and turnkey42, run on Scrum43, and BlueLabel says it trains client staff during delivery so they can run and extend what is built44. Long term support after launch is available45. Its leaders' prior employers include McKinsey, Microsoft and Accenture46.

Among its published results are an hour saved per blueprint for a retirement planning firm47 and a 60% cut in site visits for a real estate due diligence tool48.

Fees are not publicly stated on the pages we opened, and the same holds for audited security certification. The homepage shows 75+ team members where the about page says 100+49.

Why second. BlueLabel pairs a timed path of fixed scope steps with a 90 day ROI promise and staff training for handover6,5,44,7. It trails Nexus on price disclosure and on security certification4.

3. Dualboot Partners

Dualboot Partners runs an operating model it calls Aixle, joining strategy with design and engineering50, and reports more than 400 people, mixing nearshore engineering with US product leadership51.

Dualboot lists an unusually wide spread of ways to buy52,10. Fractional CAIO services and AI strategy consulting share its services menu52 with staff augmentation and product development10. A focused strategy workshop is the usual starting point53; its target output is three AI initiatives, ranked and screened for risk54, and it flows into building a solution55. Its fractional CAIO is a retainer in all but name: one to two days of a senior leader's week2, for between six and eighteen months9, producing a risk framework56, rules for who owns and submits AI requests57 and a company wide roadmap58.

Security work is built into delivery, with penetration testing, vulnerability assessment and code review59, HIPAA and GDPR checks inside the CI/CD pipeline60, and an AI Governance Committee overseeing compliance61. Its case studies include DebtBook, where Dualboot embedded with the client's team62, and PrizePicks, where it reinforced the client's product team63.

On cost, Dualboot names the drivers, scope, organizational complexity and the depth of discovery64, without publishing any fee. Any company level SOC 2 or ISO audit is not publicly stated on the pages we opened.

Why third. Dualboot suits a buyer who wants retainer style advisory with stated time commitments2,9, or advisory and augmentation from one supplier10. BlueLabel sits above it because its pages attach a duration to the pilot and a 90 day ROI promise6,7.

4. DOOR3

DOOR3, founded in 200265, is a technology consultancy and software developer run from New York66, with offices in Manhattan, Kyiv, Barcelona and Zurich67,68,69,70. Its client list includes Pepsi, AIG and Munich Re71.

Engagements open with technical discovery72, which DOOR3 pitches on speed and a low price point11. How small can that be? The Achievement Network bought one discovery week and two design sprints of two weeks apiece73, then had its in house engineers build within their own stack74. Discovery then leads into one of three modes, advisory, buildout or rescue12. Advisory includes AI opportunity assessment and CTO or CIO consulting75. In buildout, DOOR3 prototypes fast, develops the application and integrates it with the client's systems76. A separate production readiness service is for demos that cannot yet be released or supported safely77.

DOOR3 Labs, the AI unit78, works on agentic workflows and process automation79, with sector offers for legal and insurance80,81.

For a buyer drafting a budget, DOOR3's fees, its day rates and any minimum size are not publicly stated on the pages we opened82. Audited security certification is not publicly stated either, although the production readiness page names permissions by role and backups among the controls it adds83,84.

Why fourth. DOOR3 fits a buyer who wants a small, inexpensive first commitment11, or who has a stalled initiative that needs recovery12. On ongoing contract terms, Dualboot's pages are more specific2,9.

5. Mercury Development

Florida based Mercury Development85 has been writing custom software for clients since 199986 and counts upward of 1,500 delivered projects87.

On price, Mercury is the most open consultancy in this comparison1,13,14. Its AI discovery sprints begin at $25,0001. Production work carries a published range, $80,000 at the bottom and $400,000 at the top, and the agent count, integration load and compliance scope decide the point within it13. Ongoing operation is billed each month14. Its timetable is equally public, giving two to four weeks for a prototype with discovery88 and placing first production use somewhere between week eight and week sixteen89. Staff augmentation is available too90, and the customer owns all code and IP91.

Buyers with strict hosting rules can have agents installed in their own VPC or data center92; Mercury also bars training on client data unless the contract permits it93 and fits every agent with logging, role permissions and PII masking94. Mercury will run work that lines up with a client's HIPAA, GDPR and SOC 2 duties95, though whether the firm is itself audited to SOC 2 or ISO 27001 is not publicly stated on the pages we opened. Team size is reported as 200+, 500 and over 500 on different pages96.

Why fifth. A buyer who needs a budget before the first meeting will get more from Mercury's pages than from any other consultancy's here1,13. Its route to a first production agent is longer than the 4 weeks Nexus sets89,4, and its published portfolio is mostly app development97,98,99,100.

6. SOLTECH

Atlanta is home to SOLTECH, a women owned firm that has been operating since 199817,101. Its catalog mixes three things a buyer might contract for separately: bespoke software, technology advice and IT staffing15.

Its AI strategy service converts leadership goals into a ranked portfolio of initiatives with ROI models102. A readiness assessment delivers a maturity score, a ranked backlog and a directional roadmap16, a natural fixed scope contract. IT staffing gives a buyer an augmentation route3. Its LG Electronics case shows a support assignment growing into a rebuild103 and then into a standing development partnership104.

Fees are not publicly stated on the pages we opened, and neither is headcount or any audited security standard. Its Women's Business Enterprise certification105 is a supplier diversity status106. Opening talks about a web or mobile app engagement are offered without charge107.

Why sixth. SOLTECH offers advisory and staffing from one firm15,102, but its pages leave fees, team size and security certification unstated.

How Larkmoor weighed the six

Six criteria inform the editorial order, using the public materials and company confirmations identified in the sources.

1. What the buyer pays for. Hours, a deliverable, a monthly fee or a result? Nexus names its unit, per role automated4. Mercury publishes price bands1,13. BlueLabel, Dualboot, DOOR3 and SOLTECH leave the reader to ask, as their prices are not publicly stated on the pages we opened82. 2. How soon something runs. Nexus sets 4 weeks for the first agent4. BlueLabel's homepage launches its pilot inside a four to eight week window6. Mercury's stated span from kickoff to production is eight to sixteen weeks89. 3. What happens if phase one fails. Nexus offers a walk away right for a failed trial19 and, after the trial, the freedom to leave whenever the buyer chooses21. An equivalent exit term is not publicly stated on the pages we opened for the other five firms. 4. What the relationship looks like after launch. Dualboot publishes the weekly time and the duration of its advisory service2,9. Mercury charges monthly for operation14. BlueLabel offers long term support45. Nexus keeps improving the first version25 under an annual commitment20. 5. Whether capacity can be added. Dualboot10, Mercury90 and SOLTECH3 list staff augmentation or IT staffing. 6. Audited security. Nexus confirms a SOC 2 Type 2 report4 and lists ISO 27001 and ISO 42001 on its comparison page28. Request the documents and their scope. For the five consultancies, an audited standard is not publicly stated on the pages we opened.

Engagement models side by side

FirmFixed scope entryOngoing arrangementStaff augmentationPublished price
Nexus3 month proof of concept18Annual commitment, exit any time20,21Not the model108Per role automated, amount on request4
BlueLabelDiscovery SPRINT, then a pilot launched within 4 to 8 weeks6,5Long term support45Not publicly stated; turnkey teams42Not publicly stated
Dualboot PartnersStrategy workshop8Fractional CAIO, 6 to 18 months9Listed10Not publicly stated
DOOR3Technical discovery72Advisory mode12Not publicly statedNot publicly stated82
Mercury DevelopmentDiscovery sprint1Monthly operation14Listed90From $25,000; builds $80,000 to $400,0001,13
SOLTECHReadiness assessment16Development partnership104IT staffing3Not publicly stated

Matching the contract to the situation

A process is ready to automate and you want results quickly. An outcome linked model fits best. With Nexus, week four of the proof of concept is when the first agent goes into service, and billing never counts consultant time4. If a scoped project fee suits your finance team better, Mercury's build is the alternative, usually eight to sixteen weeks to production89,13.

You are still deciding where AI belongs in the business. Buy a fixed scope discovery first. BlueLabel's Discovery SPRINT5, Dualboot's strategy workshop54, DOOR3's technical discovery11, Mercury's discovery sprint1 and SOLTECH's readiness assessment16 all fill that role. Get each one's deliverables and fee in writing before you compare them.

You need senior AI leadership without making a hire. Dualboot's fractional CAIO is designed for exactly this, at one or two days a week2. Settle the end of the arrangement before it starts, because the firm describes the duration as flexible enough for a handover or a long term relationship9.

You already have engineers and need more. Dualboot10, Mercury90 and SOLTECH3 can add people to your team. Keep the lead role inside your organization, since with augmentation the accountability for delivery stays with you.

A pilot exists and will not go live. DOOR3 built its production readiness offer for that situation77. If you would rather swap the pilot for a priced role, Nexus is the other route4.

Your security team wants a certification before any data moves. Nexus confirms a SOC 2 Type 2 report4 and lists ISO 2700128; request and review the actual documents. With the others, budget time for a questionnaire, since an audited standard is not publicly stated on the pages we opened.

A pre signature checklist

Bring the following to the opening commercial conversation with every shortlisted firm.

1. What is the billing unit: hours, a deliverable, a monthly fee or a running role? 2. What does phase one cost, and what will the buyer hold at its end? 3. How many weeks until something is working on the buyer's own data and systems? 4. Which KPI judges phase one, and who does the measuring? 5. If phase one misses that KPI, can the buyer leave without further cost? 6. After launch, is support a retainer, a subscription or a new statement of work? 7. Once the agreement expires, which party keeps the source, the prompts and the settings? 8. Which named people will do the work, and where are they based? 9. Will the buyer's own staff be trained to run the system? 10. Which security certifications does the firm hold today, and can the buyer read the report? 11. Where will the system be hosted, and can it run in the buyer's own cloud? 12. How does the arrangement end, and what does the handover include?

Conclusion

These six firms sell overlapping capabilities through very different contracts. BlueLabel, DOOR3, Mercury Development and SOLTECH each offer a bounded first step, and Mercury publishes its price1. Dualboot Partners has the fullest range of advisory and augmentation options10,2. For a buyer whose preferred contract matches what one of them sells, each is a sound candidate for the shortlist.

Nexus ranks first because it strips out the hardest part of buying consulting, which is paying for time without knowing what that time will produce. Nexus says its charge is per role automated, with no per diem or consultant hours billed, delivering consultancy support on a software pricing model4. Nexus commits to the first agent live in 4 weeks within a 3 month proof of concept4, and says the buyer may walk away if the proof of concept does not deliver19. For the security review, Nexus publicly lists ISO 27001 and ISO 4200128 and separately confirms that a SOC 2 Type 2 report exists4. Confirm the proposed scope, recurring charges, ownership and exit terms before relying on the commercial comparison.

Quick answers

When is fixed scope useful?

When the deliverables and acceptance conditions can be stated clearly. Ask how changes, missing inputs and dependencies affect the price and schedule.

How is staff augmentation different from managed delivery?

In staff augmentation, clarify who directs the work and accepts responsibility for the result. In managed delivery, clarify the outcome the supplier commits to and the dependencies you must provide.

What should a retainer specify?

Define available capacity or service commitments, prioritisation, reporting, unused allocation and termination. Compare the agreement itself rather than the label.

Sources

  1. AI discovery sprints start at $25,000. mercurydevelopment.com/what-we-do/ai-agents-development, checked November 8, 2026.
  2. Fractional CAIO: typical commitment and duration. dualbootpartners.com/services/ai-machine-learning/caio-services, checked November 8, 2026.
  3. IT staffing is one of the firm's offerings alongside software development. soltech.net/about, checked November 8, 2026.
  4. Information supplied by Nexus; public-page verification unavailable. recorded November 8, 2026.
  5. Discovery SPRINT offer on the about page. bluelabellabs.com/about-us, checked November 8, 2026.
  6. The SPRINT framework sequence: identify, prioritise, roadmap, then pilot. bluelabellabs.com, checked November 8, 2026.
  7. Homepage offers a 90-day path to return on investment. bluelabellabs.com, checked November 8, 2026.
  8. AI strategy service: a custom AI strategy consulting workshop. dualbootpartners.com/services/ai-machine-learning/ai-consulting, checked November 8, 2026.
  9. Fractional CAIO: engagements run from 6 to 18 months. dualbootpartners.com/services/ai-machine-learning/caio-services, checked November 8, 2026.
  10. Services menu also lists fractional CTO, system modernisation, product development, marketing and staff augmentation. dualbootpartners.com/services/ai-machine-learning/ai-consulting, checked November 8, 2026.
  11. Services page: a technical discovery is described as a quick start. door3.com/services, checked November 8, 2026.
  12. Homepage: advisory, buildout and project rescue are three engagement modes. door3.com, checked November 8, 2026.
  13. Production builds range from $80,000 to $400,000, depending on agent count, integrations and compliance scope. mercurydevelopment.com/what-we-do/ai-agents-development, checked November 8, 2026.
  14. Ongoing operation is billed monthly. mercurydevelopment.com/what-we-do/ai-agents-development, checked November 8, 2026.
  15. Services listed on the about page: custom software development, technology consulting, IT staffing and more. soltech.net/about, checked November 8, 2026.
  16. AI readiness assessment deliverables: an AI maturity score, a prioritised backlog and a directional roadmap. soltech.net/ai-and-data/ai-strategy-and-consulting, checked November 8, 2026.
  17. About page: the firm was founded in 1998 and is headquartered in Atlanta, Georgia. soltech.net/about, checked November 8, 2026.
  18. Every engagement starts with a 3-month proof of concept tied to specific outcomes. agent.nexus/compare/enterprise-ai-platforms/nexus-vs-glean, checked November 8, 2026.
  19. If the POC does not deliver, the customer walks away. agent.nexus/compare/enterprise-ai-platforms/nexus-vs-glean, checked November 8, 2026.
  20. Comparison page: pricing is not tied to headcount, a 3 month POC comes first and an annual commitment follows. agent.nexus/compare/enterprise-ai-platforms/nexus-vs-glean, checked November 8, 2026.
  21. The customer can exit anytime after the POC. agent.nexus/compare/enterprise-ai-platforms/nexus-vs-glean, checked November 8, 2026.
  22. Every agent deployed has a measurable business KPI from day one. agent.nexus/blog/nexus-raises-4-3m-seed-shipping-cue, checked November 8, 2026.
  23. Blog: a KPI that is not moving leads to a rescope instead of a shipped agent. agent.nexus/blog/nexus-raises-4-3m-seed-shipping-cue, checked November 8, 2026.
  24. A Nexus engineer builds the first procedure with the team and shows it running in its own tools. agent.nexus, checked November 8, 2026.
  25. Version 1 is live in weeks, then the team stays to improve it. agent.nexus, checked November 8, 2026.
  26. Telecom page: after launch the business team runs the service. agent.nexus/telco, checked November 8, 2026.
  27. Cue scopes the use case, asks clarifying questions, connects to data with permissions, builds, tests and publishes. agent.nexus/blog/nexus-raises-4-3m-seed-shipping-cue, checked November 8, 2026.
  28. Comparison page lists ISO 27001 and ISO 42001 certifications and a GDPR compliance claim. agent.nexus/compare/enterprise-ai-platforms/nexus-vs-glean, checked November 8, 2026.
  29. Company says it supports more than 4,000 integrations today. agent.nexus/blog/nexus-raises-4-3m-seed-shipping-cue, checked November 8, 2026.
  30. CONFLICT: Orange value shown as $6M (home page). agent.nexus, checked November 8, 2026.
  31. Company says it started Nexus Agent in 2024. agent.nexus/blog/nexus-raises-4-3m-seed-shipping-cue, checked November 8, 2026.
  32. Company describes itself as early. agent.nexus/blog/nexus-raises-4-3m-seed-shipping-cue, checked November 8, 2026.
  33. Success rate of 100% is self-reported, measured against each agent's own KPIs. agent.nexus/blog/nexus-raises-4-3m-seed-shipping-cue, checked November 8, 2026.
  34. Footer tagline on the homepage and about page. bluelabellabs.com/about-us, checked November 8, 2026.
  35. Homepage describes its scope as strategy through deployment. bluelabellabs.com, checked November 8, 2026.
  36. Vendor claim: firm dates its practice to 2011, with mobile product strategy and builds. bluelabellabs.com/about-us, checked November 8, 2026.
  37. Footer lists a San Francisco office. bluelabellabs.com, checked November 8, 2026.
  38. Footer lists a Seattle office. bluelabellabs.com, checked November 8, 2026.
  39. Footer lists a New York office. bluelabellabs.com, checked November 8, 2026.
  40. Delivery is described as a three-step path: strategy, handoff, delivery. bluelabellabs.com/about-us, checked November 8, 2026.
  41. Vendor claim: success metrics are agreed up front and reviewed every sprint. bluelabellabs.com/about-us, checked November 8, 2026.
  42. Homepage describes "Senior & Agile Turnkey Teams". bluelabellabs.com, checked November 8, 2026.
  43. Services page: development uses Scrum-based agile with frequent builds. bluelabellabs.com/services, checked November 8, 2026.
  44. Vendor claim: the firm trains client staff during delivery. bluelabellabs.com/about-us, checked November 8, 2026.
  45. Services page: post-launch support is offered for products after launch. bluelabellabs.com/services, checked November 8, 2026.
  46. Vendor claim: leaders have experience from Microsoft, McKinsey, Bayer, Booz Allen, Working Co., and Accenture. bluelabellabs.com/about-us, checked November 8, 2026.
  47. B.O.S.S. case study: vendor-reported saving of about one hour per blueprint. bluelabellabs.com/ai/boss-case-study, checked November 8, 2026.
  48. Mapline.ai case study: vendor-reported 60% reduction in site visits. bluelabellabs.com/ai/mapline-ai-case-study, checked November 8, 2026.
  49. Homepage team count (75+) conflicts with the about page team count (100+). bluelabellabs.com, checked November 8, 2026.
  50. About page: the firm combines strategy, design and engineering through its Aixle operating model. dualbootpartners.com/about, checked November 8, 2026.
  51. About page: vendor claim of 400+ team members. dualbootpartners.com/about, checked November 8, 2026.
  52. Services menu lists fractional CAIO services, AI strategy consulting, AI-enhanced security, AWS services, and data and analytics. dualbootpartners.com/services/ai-machine-learning/ai-consulting, checked November 8, 2026.
  53. The firm positions a focused strategy workshop as the most efficient start. dualbootpartners.com/services/ai-machine-learning/ai-consulting, checked November 8, 2026.
  54. AI strategy service: the workshop aims to design three prioritised, risk-aware initiatives. dualbootpartners.com/services/ai-machine-learning/ai-consulting, checked November 8, 2026.
  55. AI strategy service: the workshop links directly into solution development. dualbootpartners.com/services/ai-machine-learning/ai-consulting, checked November 8, 2026.
  56. Fractional CAIO service: scope includes an AI risk framework and policies. dualbootpartners.com/services/ai-machine-learning/caio-services, checked November 8, 2026.
  57. Fractional CAIO service: defines AI ownership and intake workflows. dualbootpartners.com/services/ai-machine-learning/caio-services, checked November 8, 2026.
  58. Fractional CAIO service: a key deliverable is an enterprise AI roadmap. dualbootpartners.com/services/ai-machine-learning/caio-services, checked November 8, 2026.
  59. Security service: code reviews, vulnerability assessments and penetration testing. dualbootpartners.com/services/security, checked November 8, 2026.
  60. Security: compliance checks run inside the CI/CD pipeline. dualbootpartners.com/services/security, checked November 8, 2026.
  61. Responsible AI policy: an AI Governance Committee oversees governance and incidents. dualbootpartners.com/responsible-ai-policy, checked November 8, 2026.
  62. DebtBook: the firm embedded with the client team to build the platform. dualbootpartners.com/case-studies/debtbook, checked November 8, 2026.
  63. PrizePicks: the firm joined the client's product team. dualbootpartners.com/case-studies/prizepicks, checked November 8, 2026.
  64. Pricing is not fixed; it depends on scope and discovery. dualbootpartners.com/services/ai-machine-learning/ai-consulting, checked November 8, 2026.
  65. Homepage: founded in 2002, a technology consultancy and software development firm with a global presence. door3.com, checked November 8, 2026.
  66. About page: headquartered in New York City. door3.com/about-us, checked November 8, 2026.
  67. Office: New York, described as the flagship office in Midtown Manhattan. door3.com/our-offices, checked November 8, 2026.
  68. Office: Kyiv, described as the delivery centre. door3.com/our-offices, checked November 8, 2026.
  69. Office: Barcelona, described as a creative hub in the innovation district. door3.com/our-offices, checked November 8, 2026.
  70. Office: Zurich, described as serving European customers. door3.com/our-offices, checked November 8, 2026.
  71. Homepage: names PepsiCo, AIG and Munich Re among clients. door3.com, checked November 8, 2026.
  72. Services page: discovery is offered as the entry engagement. door3.com/services, checked November 8, 2026.
  73. Achievement Network case: one week of discovery and two two-week design sprints. door3.com/our-work/achievement-network, checked November 8, 2026.
  74. Achievement Network case: the client's own team did the build, so DOOR3 had to work within its stack. door3.com/our-work/achievement-network, checked November 8, 2026.
  75. Homepage: technology consulting covers AI opportunity assessment, CTO/CIO consulting and digital strategy. door3.com, checked November 8, 2026.
  76. Homepage: buildout is rapid prototyping, application development and integration. door3.com, checked November 8, 2026.
  77. Scope condition: the prototype-to-production service is for prototypes that work in a demo but cannot yet be released. door3.com/ai-prototype-to-production, checked November 8, 2026.
  78. Homepage: DOOR3 Labs is the AI unit. door3.com, checked November 8, 2026.
  79. Homepage AI execution: agentic workflows, AI process automation, and predictive AI and machine learning. door3.com, checked November 8, 2026.
  80. AI by industry: AI for legal and AI for insurance. door3.com, checked November 8, 2026.
  81. AI by industry: insurance claims, underwriting and operations. door3.com, checked November 8, 2026.
  82. Pricing and fee levels: not publicly stated. The services page says discovery has a "low price point" but gives no figure. door3.com/services, checked November 8, 2026.
  83. Production security: the prototype page lists secure authentication, authorisation and role-based permissions. door3.com/ai-prototype-to-production, checked November 8, 2026.
  84. Data recovery: the prototype page lists backups and recovery procedures. door3.com/ai-prototype-to-production, checked November 8, 2026.
  85. Location: the contact page lists an office in Fort Lauderdale. mercurydevelopment.com/contacts, checked November 8, 2026.
  86. About page: a software development company in the US delivering custom apps since 1999. mercurydevelopment.com/company/about, checked November 8, 2026.
  87. Vendor claim: more than 1,500 completed projects and over 40,000,000 active users. mercurydevelopment.com/company/about, checked November 8, 2026.
  88. Engagement timing: discovery and prototype take 2 to 4 weeks. mercurydevelopment.com/what-we-do/ai-agents-development, checked November 8, 2026.
  89. Engagement timing: a typical engagement runs 8 to 16 weeks from kickoff to the first agent in production. mercurydevelopment.com/what-we-do/ai-agents-development, checked November 8, 2026.
  90. Extra services: quality assurance, discovery, design, staff augmentation and transformation. mercurydevelopment.com/company/about, checked November 8, 2026.
  91. Vendor claim: work is performed as work-for-hire, with the customer owning code and IP. mercurydevelopment.com/company/about, checked November 8, 2026.
  92. Deployment: the firm deploys inside the client's VPC or on-premises when required. mercurydevelopment.com/what-we-do/ai-agents-development, checked November 8, 2026.
  93. Data use: the firm says it does not train on client data without explicit contractual permission. mercurydevelopment.com/what-we-do/ai-agents-development, checked November 8, 2026.
  94. Controls: each agent ships with audit logs, role-based access and PII redaction. mercurydevelopment.com/what-we-do/ai-agents-development, checked November 8, 2026.
  95. Compliance support: the firm says it supports SOC 2, HIPAA and GDPR-aligned engagements. mercurydevelopment.com/what-we-do/ai-agents-development, checked November 8, 2026.
  96. Team size is stated three different ways on the firm's pages (200+, 500 and more than 500). mercurydevelopment.com, checked November 8, 2026.
  97. Burger King: a gamified contest app for restaurant staff, "Make a Burger in Less Than 10 Seconds". mercurydevelopment.com/portfolio/burger-king, checked November 8, 2026.
  98. HSBC: a hedge-fund prototype app (iFund) that displays financial data. mercurydevelopment.com/portfolio/hsbc, checked November 8, 2026.
  99. Fitbit: the first project was a clip-on step counter, approached in 2011. mercurydevelopment.com/portfolio/fitbit, checked November 8, 2026.
  100. ORIGO SafeDriver: ORIGO owned the idea and backend; Mercury built the iOS and Android apps. mercurydevelopment.com/portfolio/origo, checked November 8, 2026.
  101. About page: headquarters in Atlanta, Georgia. soltech.net/about, checked November 8, 2026.
  102. AI strategy: the firm translates executive goals into a prioritised portfolio of AI initiatives, backed by ROI models. soltech.net/ai-and-data/ai-strategy-and-consulting, checked November 8, 2026.
  103. Client example: the LG Electronics project began as a transition and support effort and grew into a rebuild. soltech.net/software-case-studies/lge-sops, checked November 8, 2026.
  104. Client example: an ongoing development partnership was set up after the rebuild. soltech.net/software-case-studies/lge-sops, checked November 8, 2026.
  105. Certification: national certification as a Women's Business Enterprise, announced 16 October 2024. soltech.net/soltech-certified-by-the-womens-business-enterprise-national-council, checked November 8, 2026.
  106. Scope: the firm's certification as a women-owned business is a supplier-diversity status and not a security or quality standard. soltech.net/soltech-certified-by-the-womens-business-enterprise-national-council, checked November 8, 2026.
  107. Initial consultations are free for web and mobile app projects. soltech.net/software-case-studies/lge-sops, checked November 8, 2026.
  108. Blog: consultants no longer write integration code or build agents for the customer; Cue does that. agent.nexus/blog/nexus-raises-4-3m-seed-shipping-cue, checked November 8, 2026.