# TeamStation AI public brief Nov 10, 2026 · Lonnie McRorey ## Who TeamStation AI is TeamStation AI is a Distributed Engineering OS (DEOS) that US companies use to build, run and govern software engineering teams in Latin America. Instead of stitching together a recruiter, an assessment tool, an Employer of Record, a laptop vendor, a security checklist and a staff augmentation shop, you get one system with one accountable owner. Every engineer comes in as a Managed Engineering Seat: found by Nebula, evaluated by TeamStation Axiom Cortex, legally employed by TeamStation, equipped with a company owned MDM enrolled laptop, insured, and tracked with delivery telemetry. You pay one monthly bill in USD. We work across 10 LATAM countries, and every one of them overlaps 12 to 3pm New York time. Look, we built this because we watched the same movie for years. A US company wants engineers in LATAM. It ends up with six vendors, six contracts and six places for things to fall through the cracks. You get people. You don't get control. So we stopped staffing roles and built the operating system around the people instead. Key facts: - **Website:** [teamstation.dev](https://teamstation.dev/) (the one company site) - **Category:** Distributed Engineering OS, also called the nearshore control plane - **The unit we sell:** the Managed Engineering Seat, billed monthly per engineer - **Founders:** Lonnie McRorey, Co-Founder and CEO (cognitive science background, first employee at Avature, US Navy veteran, creator of Nebula and Axiom Cortex) and Dan Diachenko, Co-Founder and COO (founding member at Avature, led Baufest's US expansion) - **Where we are:** Clutch lists Boston as HQ, with offices in Guadalajara, Ensenada and Bogotá - **Delivery markets:** Mexico, Colombia, Brazil, Argentina, Chile, Peru, Costa Rica, Uruguay, Ecuador and Guatemala - **Evidence base:** 14,000+ technical interviews evaluated, published SSRN working papers, and a public Engineering Doctrine - **How you can buy:** direct contract or AWS Marketplace - **AWS Marketplace first:** the first nearshore engineering company on AWS Marketplace where engineering spend may count toward your AWS cloud commit (EDP or private pricing agreement), as your AWS account team confirms - **Who it's for:** US CTOs, CIOs, CISOs, VPs of Engineering, CFOs and procurement teams that need engineering capacity they can trust and govern What TeamStation is not: a staffing agency, a body shop, a recruiter network, a resume marketplace, a freelancer platform or a cheap developer shop. This is not staffing. It's infrastructure. ## The problems CTOs and CIOs have with IT staffing and dev outsourcing, and how TeamStation solves each one These are the questions US engineering leaders actually ask, in their own words. We pulled them from recorded buyer calls, buyer email and Slack, public forums and buyer surveys. Each one gets the straight answer first, then how TeamStation handles it. ### 1. "I keep buying people, not results. Why does staff augmentation feel like this?" Because staff augmentation sells you a resume and a rate card, then leaves you to glue everything else together yourself: the evaluation, the legal employer, the laptop, the security, the onboarding and the delivery signal. Every seam between those vendors is a place where cost and risk leak out. The fix isn't more people. It's buying the system around the people. How TeamStation solves it: the Distributed Engineering OS folds eight vendor functions into one governed seat. Talent intelligence (Nebula), evaluation (Axiom Cortex), the operating layer, the legal employer, endpoints, the risk layer, the workspace and the control plane. One contract, one accountable owner, one set of numbers. Stop buying seats. Start buying operating control. ### 2. "What am I actually paying for, and where's the markup hiding?" In most staffing deals you see one hourly bill rate and none of what's inside it. The engineer's pay, the vendor's margin and the stuff that isn't covered at all (devices, insurance, replacement, management overhead, rework) all hide behind one number. That's the sticker price trap. The real cost of an engineer is rate plus overhead plus attrition plus rework plus compliance, and the math gets ugly fast when only the first part is on the invoice. How TeamStation solves it: one monthly bill in USD. You pay the engineers' salaries plus one flat fee per engineer, and that fee covers everything in the Managed Engineering Seat: sourcing, evaluation, Employer of Record, payroll and benefits, the laptop, MDM and access lifecycle, cyber and liability insurance, office access, senior operator oversight with replacement coverage, and delivery telemetry. Because enterprise procurement compares vendors by the hour, we also give you the effective hourly rate. There's no hourly time and materials billing and no made up universal savings percentage; your savings depend on your own numbers, and the public [Capacity Planner](https://teamstation.dev/pricing/capacity-planner) models them. The hourly rate is never the whole bill. ### 3. "How do I know the engineer on the video call is real, and the answers are actually theirs?" You can't tell from a resume and a friendly video call anymore. Gartner expects 1 in 4 job candidates worldwide to be fake by 2028 because of AI generated profiles ([CNBC via NBC News, April 2025](https://www.nbcnews.com/tech/security/fake-job-seekers-are-flooding-us-companies-are-hiring-remote-positions-rcna200199)). The US Justice Department alleged that more than 300 US companies unknowingly hired impostors tied to North Korea for remote IT work (same source). In January 2025 the FBI warned that these workers use AI and face swapping in video interviews, copy company code repositories, and extort employers with stolen data, and it told companies to verify that third party staffing firms run thorough hiring checks ([FBI IC3 alert, Jan 23, 2025](https://www.ic3.gov/PSA/2025/PSA250123)). Interview fraud is now a security problem, not just a hiring problem. How TeamStation solves it: every TeamStation Axiom Cortex interview is recorded and transcribed with timestamps, and integrity checks run first: liveness, duplicate and scripting checks. Human reasoning has natural variance. Scripted or AI polished answers show unnaturally smooth patterns, and those get flagged to a human reviewer, never auto rejected. Only the candidate's own words count. Then the engineer becomes our legal employee in a known country, working on a company owned, MDM enrolled laptop. Identity, contract, device and evaluation live in one record, and you see the evidence before you decide. ### 4. "The senior I interviewed isn't who showed up, or isn't really senior. How do I stop the bait and switch?" Bait and switch happens two ways. The person you interviewed isn't the person placed, or "senior" turns out to be a title, not a skill. Titles inflate, resumes get polished by AI, and an engineer can ship fast with AI tools for months and still break the first time the system does something new. The fix is to tie the evidence to the person and to measure reasoning, not vocabulary. How TeamStation solves it: the evidence you review belongs to the engineer we hire, because evaluation, employment and the device sit in the same system. TeamStation Axiom Cortex scores each recorded answer on its own against a blueprint built from your job, across five Behavioral Axioms: accuracy, mental model, procedural knowledge, clarity and cognitive load. It also checks whether confidence matches knowledge, since confident and wrong is the most dangerous hire there is. We separate syntax seniors (they know the keystrokes) from cognitive seniors (they can hold the real system in their head). Our 2026 research on Axiom Cortex v3, run on TeamStation's own hiring data, reports false positives falling from about 42% to 8% compared with conventional screening. And our bar is blunt: we'd rather reject five good engineers than hire one bad one. ### 5. "Who legally employs the engineer, and what labor law risk am I carrying?" There are only three doors. Hire contractors, open your own entity, or use an Employer of Record. A contractor who works full days inside your Jira, your Slack and your standups can be treated as an employee by local labor authorities, and that risk lands on you. Some countries have tightened the rules on outsourcing people: Mexico's 2021 reform prohibited subcontracting of personnel except registered specialized services, took away tax deductions for non compliant arrangements, and added criminal penalties for sham schemes ([BDO summary of the reform](https://www.bdo.com/insights/tax/mexico-enacts-far-reaching-labor-reform-that-addresses-outsourcing-arrangements)). Calling a full time engineer a contractor is like calling your house a tent because you went camping once. The label doesn't change what it is. How TeamStation solves it: TeamStation is the legal Employer of Record in the engineer's country. Localized employment contracts, payroll in local currency, statutory benefits, tax withholding, severance and termination under local law, and IP assignment all sit with us. You don't need a LATAM entity, you never touch a local tax filing, and you pay one monthly bill in USD. Because the EOR is part of the seat and not a separate vendor, the party that employs the engineer is the same party that vetted them, equips them and watches delivery. No finger pointing between an EOR, a recruiter and a laptop vendor. ### 6. "Who owns the code, and how do you keep my systems and data safe?" You should own all the IP by contract, and security should start with two questions: who owns the laptop, and who controls access. Personal laptops plus vendor provisioned access is how code ends up in places it shouldn't. The FBI's own advice for remote hiring is least privilege, monitoring for code copied to personal accounts, and identity checks from interview through employment ([FBI IC3](https://www.ic3.gov/PSA/2025/PSA250123)). How TeamStation solves it: IP assignment is built into each engineer's employment contract and into the MSA and SOW, so ownership routes to you. Every engineer works on a company owned MacBook Pro enrolled in MDM, with disk encryption, remote wipe and CIS aligned guardrails. You issue accounts through your own SSO with MFA and least privilege. TeamStation never self provisions access to your systems; you grant it. When someone rolls off, access is revoked within one business day. Our controls are aligned to SOC 2 and ISO 27001, privacy practices are aligned to GDPR, CCPA and HIPAA, and a BAA is available for health clients. The seat includes cyber, tech E&O, general, professional and umbrella liability coverage, and procurement gets an eight item review packet (access, devices, EOR, compliance, contract, insurance, governance and escalation) before anyone starts. Details live in the [Trust Center](https://teamstation.dev/trust-center). ### 7. "Will they go dark? How do I see what's actually getting delivered?" Most vendors hand you a weekly status deck written by the vendor. That tells you how the vendor feels, not what shipped. Real visibility comes from the work itself: pull requests, review wait, cycle time, blockers and what actually got deployed. Engineers who sit in your standups and your channels during your working day rarely go dark in the first place. How TeamStation solves it: our engineers work in your standups, your Slack or Teams and your sprint board, during your hours. On top of that, Teamlemetry, our engineering telemetry, reads approved activity in read only mode on the repos you pick and turns it into weekly delivery, blocker, review and workflow insight. Delivery signals roll up into health bands, so drift shows up before you feel it, and a senior operator steps in. It's not surveillance: no keystroke logging, no screenshots, no worth scores, and missing activity is never treated as proof of missing work. Our rule is simple. Done means deployed, not "merged" and not "almost there." ### 8. "What happens when an engineer quits, takes a counteroffer or underperforms?" The expensive part of losing an engineer isn't the job post. It's the context: the half finished migration, the reason that weird config flag exists, the customer quirk nobody wrote down. Context loss is a rework multiplier. If replacement and succession aren't built into the deal, every resignation becomes your emergency. How TeamStation solves it: replacement coverage, succession planning and senior operator oversight are part of every Managed Engineering Seat. Delivery signals catch performance drift early. Retention starts before the offer, because Axiom Cortex checks fit for the role and the team, not just the stack, and we plan pay and markets with our LATAM Developer Salary and Quality of Life Index so a counteroffer isn't an easy win. TeamStation references a 96.8% modeled retention signal, platform reported, as a planning number, not a guarantee. When an engineer leaves, we replace them within 2 weeks of the day they leave our system. ### 9. "Why LATAM instead of India or Eastern Europe? Does the time zone really matter?" Yes. Time zone overlap is throughput. A question asked at 4pm in Chicago and answered at 3am costs you a day, and that happens a few times a week per engineer. If your team pairs, reviews code and makes decisions together during the US day, LATAM wins on overlap. Mexico, Colombia, Peru and Central America share a full New York workday. Eastern Europe shares about 2 to 3 hours. India shares almost none without night shifts. Offshore works best for well specified, async work where an overnight handoff is a feature. | Hub | Hours vs New York (winter / summer) | Shares a full 9 to 5 ET day? | | --- | --- | --- | | Mexico City, Guadalajara | 1 behind / 2 behind | Yes | | Guatemala City, San José | 1 behind / 2 behind | Yes | | Bogotá, Lima, Quito | Same / 1 behind | Yes | | São Paulo, Buenos Aires, Montevideo | 2 ahead / 1 ahead | Most of it | | Santiago | 0 to 2 ahead, by season | Most of it | | Warsaw | 6 ahead | About 3 hours | | Bucharest, Kyiv | 7 ahead | About 2 hours | | Bengaluru | 10.5 ahead / 9.5 ahead | None without night shifts | How TeamStation solves it: every TeamStation delivery market overlaps 12 to 3pm New York, and most overlap far more. On the West Coast, Mexico and Central America sit just 1 to 2 hours ahead of Pacific time. ### 10. "How are you different from an EOR platform or a staffing agency? And what if my vendor fails me?" An EOR platform handles payroll and the legal employer, and that's it. An agency or talent marketplace finds people, and that's it. A dev shop sends a bench. None of them owns the outcome, so when something breaks you're the integrator. Vendor lock-in happens when the vendor ends up holding the knowledge, the access and the people, and you can't see inside. How TeamStation solves it: we run the whole seat in one system, so there's one owner when something breaks. We compare operating models, not reputations, and the full side by side of dev shops, staff augmentation, talent networks, EOR platforms and TeamStation is in the section right after these 18 problems, "A new category: how TeamStation replaces the legacy vendor models." On lock-in: your code, docs and decisions live in your repos and your tools, you grant and revoke access yourself, the IP is assigned to you, and the evaluation evidence on every engineer is yours to see. Thinking about pulling engineers in house later? Run the math first. You'd take on securing them, devices, insurance, legal liability for damages, country labor law, EOR and taxes. We do all of that for you through the system. We replace your vendors and become your technology arm that legally employs your engineers, without the greenfield risk or the in country risk. ### 11. "How fast can a team actually start?" Faster than most buyers expect, and slower than most vendors promise. The real clock depends on how rare the role is, how fast your team reviews candidates, your security and compliance steps, and how quickly you grant access. Most delays happen at the handoffs between vendors, not inside any one step. How TeamStation solves it: we plan to a 9 day average time to squad. That's a planning target, not a promise. Hiring runs as a 10 stage pipeline with 3 interview steps, and you review evidence (recorded interview, transcript, scores) instead of a resume pile, so reviews move faster. The employment contract, the MDM enrolled laptop and access setup run in parallel instead of one after another. In one public case, a Fortune 500 ERP team on Dynamics F&O shipped its first pull request in 14 days. Once candidates are ready, getting to the final interview and the hire usually takes days, because of our intelligent system. ### 12. "Will they work inside our tools and our process?" They should, or it isn't really your team. A vendor that makes you work through its own project manager and status deck adds a translation layer, not capacity. How TeamStation solves it: yes. TeamStation engineers work inside your Slack or Microsoft Teams, your single sign on (Okta, for example), your Jira, your repos and your rituals. You grant and control access. They show up on day one with their own company owned, MDM enrolled laptop. Billing is one flat monthly seat, so you don't need hourly timesheets to get invoiced. ### 13. "Can your engineers actually work AI native and own outcomes with little supervision?" The real question isn't whether an engineer uses AI tools. Everybody does now. It's whether they can hold the system in their head while agents produce the volume, because review capacity is the new bottleneck in the SDLC. Polished syntax can hide missing understanding. How TeamStation solves it: we build agentic AI squads across a six stage delivery chain: Frame, Design, Build, Review, Verify and Run. Engineers are placed by cognitive fit, not just by stack, and the offer starts with Forward Deployed Engineers, in pods of 1 to 4. The rule inside the squad: a named human owns the merge, and agents carry the volume under review. Give AI agents bounded work, not open authority. Axiom Cortex is how we find the cognitive seniors who can run that loop. ### 14. "Which LATAM country should we hire in?" It depends on the role, the stack, the seniority and the budget, plus how much overlap you need. No single country wins everything. How TeamStation solves it: we hire and employ in 10 markets, and each has a different strength. | Country | Where it shines | | --- | --- | | Mexico | Mature nearshore market | | Colombia | Product and cloud | | Brazil | Large talent pool | | Argentina | Senior depth | | Chile | Data and fintech | | Peru | Mobile and web | | Costa Rica | Enterprise delivery | | Uruguay | Specialized senior talent | | Ecuador | Emerging cost and quality | | Guatemala | Central Time overlap | Squads can mix countries, which also spreads holiday and attrition risk. The [Capacity Planner](https://teamstation.dev/pricing/capacity-planner) and the [LATAM Developer Salary and Quality of Life Index](https://teamstation.dev/latam-developer-salary-quality-of-life) compare markets side by side. ### 15. "What do the contract and the buying process look like? Can we pilot first?" Procurement asks the same questions every time: whose paper, how long, how do we get out, where's the insurance certificate, and can we start small. Most vendors make you sign an NDA just to find out. It's a restaurant that won't show you the menu until you've ordered. How TeamStation solves it: one MSA and one SOW per team or squad, covering the full seat (roles, levels L1 to L5, country mix, start dates and the monthly seat rate). One monthly bill in USD. IP routes to you. Mutual NDA, plus confidentiality in every engineer's employment contract. A certificate of insurance on request; coverage limits are covered on a call, not on the website. The eight item procurement packet arrives before launch. You can start with a Lean Pilot Squad and scale under the same MSA. Terms usually start at a 6 month minimum, and the details live in the MSA and SOW. You can buy direct, or through AWS Marketplace ("Nearshore Engineering Services for AWS"). TeamStation is the first nearshore engineering company on AWS Marketplace where engineering spend may count toward your AWS cloud commit (EDP or private pricing agreement), and your AWS account team confirms whether it counts. ### 16. "How do holidays, PTO and on-call work across countries?" Engineers follow the public holidays and statutory vacation of the country where they're employed, not US holidays. Each country has its own calendar, and Colombia, for one, moves many holidays to a Monday. None of it is hard; it just has to be on the sprint plan before the sprint. On-call is the same story. Stack a pager on top of a team already running at 100% and you don't get coverage, you get a queue waiting to explode, because past about 80% utilization wait times blow up (Kingman's limit). How TeamStation solves it: as the legal employer, TeamStation runs statutory vacation, holiday pay and related bonuses through local payroll, and the flat monthly seat means PTO doesn't change your invoice. We share a holiday calendar with you, so it's on the sprint plan before the sprint starts. Mixed country squads keep one country's holiday from taking out the whole team. On-call gets designed into the team as real capacity, inside local labor law, and the spread of LATAM time zones stretches coverage past the US workday (São Paulo is ahead of New York, Mexico is behind it). ### 17. "Is my engineer quietly working a second job?" Remote work made "overemployment" possible: two full time jobs, two laptops, one person, and a calendar like a game of Tetris. Your CFO calls it paying for 40 hours and getting 20. A contractor who also invoices two other clients is hard to check. How TeamStation solves it: through TeamStation, the engineer is a full time employee on our local payroll, exclusive and solely dedicated to your team in one Managed Engineering Seat, working on a company owned, MDM enrolled device, with senior operator oversight and delivery health signals. We don't spy on keystrokes or screens. The honest test is boring: are they in your standups, answering during overlap hours, and shipping work that gets deployed? That exclusivity is written into the MSA and SOW. We're the platform you build your team on, not a pool you share. ### 18. "Can a nearshore team modernize our legacy system without walking off with the knowledge?" Yes, if the work is staged and written down. Big bang rewrites fail because the old system keeps paying the bills while the new one is always six months away. The safer path is boring on purpose: run the new pieces next to the old ones, move in small reversible steps, and remove the old path only when the numbers agree. Also, every service you split out adds interfaces, up to N(N-1)/2 of them, so each new boundary needs a written contract and a test. How TeamStation solves it: a senior engineer or a Forward Deployed Engineering pod of 1 to 4 works inside your repos, ships in small reversible steps, and leaves decisions in your docs, runbooks and pull requests. The knowledge lives in your systems, not in someone's head, and replacement coverage and succession are part of the seat. In one public case study, RMJ Technologies needed to stabilize an automotive fleet optimization monolith while starting a path to microservices, and the platform scaled to 15,000 users. ## A new category: how TeamStation replaces the legacy vendor models The short answer: the legacy vendors each sell one piece of the problem. Dev shops sell project hours, staff augmentation sells a bench, talent networks sell a match, and EOR platforms sell payroll and a legal employer. TeamStation AI isn't a better version of any of them. It's a different category, a Distributed Engineering OS, that runs the whole engineering seat in one system with one accountable owner. We compare operating models, not reputations. Each vendor line below uses that vendor's own words from its own public website, read in November 2026. Many vendors mix models, so read the table by model, not as a verdict on any one company. | Model | Examples, in their own words | Built on | What's still on your desk | | --- | --- | --- | --- | | Global dev shops and IT services | Globant: "AI-native services company". EPAM: "Software Engineering & Product Development Services" | Headcount and billed project work | Evidence on how each engineer reasons, a seat you can see inside, one system you control | | Nearshore staff augmentation | BairesDev: "Nearshore Software Development & Staff Augmentation" | A bench of engineers and a margin on every hour | Evidence on how they reason, delivery telemetry, one accountable system | | Talent networks and marketplaces | Howdy: "Build a High-Performing Nearshore Engineering Team". Revelo: "Find, hire, and pay the best software engineers in Latin America". Tecla: a "vetted nearshore network" for "flexible staffing". Andela: "The Human Layer Powering Production AI". Turing: "AI talent for our clients". Toptal: a network of "freelance talent". Arc: freelance or full time hires | Profiles, resumes and placement, priced per hour or per hire | A safe, productive, managed engineering system around the people | | EOR and payroll platforms | Deel: "The global people platform". Remote: "One platform for global payroll, EOR, and contractor management". Oyster: "Hire, pay, and care for talent" | Payroll and legal employment, per worker | Finding, vetting, devices, security, delivery | | Offshore incumbents | Large offshore IT services firms | Scale and labor cost | Time zone overlap and evidence before commitment | | DIY hiring | Your own recruiters, entity or EOR, and IT | Your own payroll and overhead | Every vendor and every risk, run by your team | | TeamStation AI | Distributed Engineering OS | Evidence on how engineers reason, and one governed seat billed monthly | Your product decisions and your code reviews | Some vendors bundle more than their model. Howdy says it handles recruiting, payroll, benefits, equipment, compliance and ongoing support. Tecla also offers built outcomes through Tecla Labs. Revelo, Turing and Andela now also do AI data and AI delivery work. That's fine. The question isn't who bundles the most. It's what the whole thing is built on. Why this is a new category, not a better vendor: - **Science, not resumes:** TeamStation Axiom Cortex evaluates how an engineer reasons, communicates and handles ambiguity, built on 14,000+ evaluated technical interviews. You see the recorded interview, the transcript and the scores before you commit. - **One Managed Engineering Seat:** sourcing by Nebula, evaluation, Employer of Record in country, payroll and benefits, a company owned MDM enrolled laptop, insurance, senior operator oversight and delivery telemetry. One monthly bill in USD, with the effective hourly rate shown. - **Full compliance built in:** we're the legal employer in each country, so country labor law, EOR, taxes, devices, security and liability sit with us, not with you. - **One owner, one SLA:** governance runs as a single SLA across evaluation, onboarding, devices, delivery visibility and capacity economics. When an engineer leaves, we replace them within 2 weeks of the day they leave our system. - **Built to replace the stack:** the seat replaces the recruiter, the assessment tool, the EOR, the laptop vendor, the security checklist and the staff augmentation shop. We become your technology arm that legally employs your engineers. - **Buy it like cloud:** TeamStation is the first nearshore engineering company on AWS Marketplace where engineering spend may count toward your AWS cloud commit (EDP or private pricing agreement), as your AWS account team confirms. The detail, vendor by vendor, is on teamstation.dev's 31 operating model comparison pages ([comparisons](https://teamstation.dev/comparisons)). ## How the Distributed Engineering OS and the Managed Engineering Seat work The short answer: you don't buy a person, you buy a governed seat. The Distributed Engineering OS (DEOS) is one operating layer with seven parts that hand off to each other, and the Managed Engineering Seat is what comes out the other end. 1. **Nebula finds and proves candidates.** Nebula is the LATAM talent intelligence graph, mapping 2.6M+ engineering signals by role, stack, level, country and alignment. Its rule: "Search finds URLs. Only Evidence creates truth." Every record is graded from E0 (seed only) to E4 (recruiter verified), and nothing becomes a profile fact without evidence behind it. 2. **TeamStation Axiom Cortex proves how they think.** It checks how an engineer thinks, explains work, breaks down problems, works with a team and handles pressure, before that engineer touches production work. More in the science section below. 3. **Topology places them in the delivery chain.** Engineers are matched by cognitive fit to a six stage agentic SDLC chain: Frame, Design, Build, Review, Verify and Run. 4. **The service envelope makes them ready on day one.** TeamStation is the legal Employer of Record. Company owned, MDM enrolled MacBook Pro laptops with disk encryption, remote wipe and CIS aligned guardrails. SSO, MFA, least privilege and revocation records. Controls aligned to SOC 2 and ISO 27001. Cyber, tech E&O, general, professional and umbrella liability coverage. WeWork powered office access. An eight item procurement packet before launch. 5. **Teamlemetry keeps delivery visible.** Read only engineering telemetry on the repos you pick, turned into weekly delivery, blocker, review and workflow insight. Not surveillance. 6. **Delivery governance keeps one owner accountable.** Senior operator oversight, escalation, succession and replacement coverage, under one governance model. 7. **Pricing turns it all into one seat.** Total delivery cost, not just salary, billed monthly per seat with the effective hourly rate shown. What one Managed Engineering Seat includes: - Sourcing through Nebula and evaluation through Axiom Cortex - Agentic role planning and placement in the delivery chain - Employer of Record, payroll and benefits in the engineer's country - A company owned laptop, MDM, and the access lifecycle from first login to revocation - Cyber insurance, tech E&O and liability coverage - Office access - Senior operator oversight with succession and replacement coverage - Delivery governance and telemetry - Leadership compensation planning How a team gets built: hiring runs as a 10 stage pipeline with 3 interview steps. Every candidate gets a workspace and a candidate card, so you see evidence, not a resume pile. Delivery models you can buy: - Single Managed Engineering Seats - Governed pods and squads - Agentic AI squads across the delivery chain, starting with Forward Deployed Engineers in pods of 1 to 4 - Leadership seats: Engineering Manager, Head of Engineering and a CTO path One owner, one SLA: governance runs as a single SLA across evaluation, onboarding, devices, delivery visibility and capacity economics. Partners can white label it. Recruiting agencies, MSPs and IT consultancies, fractional CTOs, PE and VC portfolios and family offices can put their own brand in front of clients while TeamStation supplies the LATAM talent graph, Axiom Cortex vetting, the pricing planner, VMS and API connections and the back office. Details at [teamstation.dev/partner](https://teamstation.dev/partner). The doctrine behind all of it: "People are rarely the problem. Systems usually are." And our rule for finished work: done means deployed. A merged pull request isn't the finish line. ## The science: Neuro-Psychometric Science and TeamStation Axiom Cortex TeamStation treats engineering capacity as a signal processing and systems problem, not a staffing problem. The science runs in three layers. Nebula maps the talent market. Axiom Cortex measures how an engineer actually reasons. The DEOS places that person in a team topology and watches delivery with telemetry. Language models only turn verified numbers into readable reports. They are never the judge. ### What "neuro-psychometric" means here It means measuring nonclinical work reasoning from evidence in what an engineer says and how they say it. It is not personality, not IQ, not emotion, and not accent, face or voice. Axiom Cortex was hardened inside TeamStation's own hiring across 14,000+ technical interviews. ### What Axiom Cortex measures Each answer is scored on five Behavioral Axioms: - **Accuracy:** is it right. - **Mental model:** does it show a real model of the system. - **Procedural knowledge:** are the steps sound. - **Clarity:** is it clear. - **Cognitive load:** how much strain it took to get there. Those roll up into four traits: Analytical Intelligence, Problem-Solving Agility, Learning Orientation and Collaborative Mindset. The Metacognitive Conviction Index checks whether confidence matches knowledge, because confident and wrong is the most dangerous hire there is. ### How an evaluation runs 1. The technical interview is recorded and transcribed with timestamps. Every question carries an ideal answer blueprint built from your job. 2. Integrity checks run first: liveness, duplicate and scripting checks. 3. Each answer is scored on its own, so one great or bad answer can't color the rest. Only the candidate's own words count, not words the interviewer supplied. 4. The answer's meaning is compared with the blueprint, so covering the right concepts in different words still scores. We measure reasoning, not recitation. 5. Behavioral answers are scored for problem, impact, agency and justification, so polished storytelling doesn't win. No evidence, no credit. 6. Scripted or AI polished answers show unnaturally smooth patterns. Those get flagged to a human reviewer, never auto rejected. 7. The Cortex Calibration Layer separates engineering reasoning from accent, grammar and culture. Hedging like "I think" isn't read as low accuracy, and a strong mental model with weaker second language clarity still gets credit. In our words: "Math does not have an accent." 8. A certified human reviewer owns the hiring decision. You see the video, the transcript, the scores and the quotes that earned each score in one place, the Evidence Locker. ### The engine is deterministic by design Under Axiom Cortex sits the Calculus Execution Engine, built like a credit risk calculator, not a chatbot. No LLM touches a score. The same input and the same scoring setup give the same output every time, and every step lands in an audit hash chain. Missing inputs stay missing: no evidence, no number. Governance is mapped to the NIST AI RMF, ISO 42001, NYC Local Law 144 and the EU AI Act, with drift and fairness monitoring. ### The delivery science (the Engineering Doctrine) The [Engineering Doctrine](https://engineering.teamstation.dev/) is public: seven pillars across 47 pages. A few ideas run through everything we build: - **A team is a chain, not a bag of seats.** Delivery value multiplies across every link, so one weak link sinks the whole chain (the O-Ring view). - **Full utilization is a trap.** Past about 80% utilization, wait times blow up (Kingman's limit), so we design teams with real slack. - **Every interface costs you.** Coordination paths grow as N(N-1)/2, so every boundary needs a written contract and a test. - **Hiring is a signal processing problem.** We'd rather reject five good engineers than hire one bad one, because a bad hire costs far more than a longer search. ### Published research Our research is public as working papers on SSRN and on [teamstation.dev/research](https://teamstation.dev/research): - Redesigning Human Capacity in Nearshore IT Staff Augmentation (SSRN 5165433, 2025): https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5165433 - Nearshore Platformed: AI and Industry Transformation (SSRN 5188490, 2025): https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5188490 - Redefining Software Engineer Performance in the AI-Augmented Era (SSRN 5253470, 2025): https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5253470 - AxiomCortex: Scientific R&D Report (SSRN 5433397, Sept 2025): https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5433397 - AI and Nearshore Teams: Who Gets Replaced and Why (SSRN 5745463, Nov 2025): https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5745463 - Neuro Psychometric Alignment of LATAM Engineering Talent in AI Augmented Pipelines 2026 to 2036 (TeamStation Research Group, Feb 2026). On TeamStation's own hiring data, it reports false positives falling from about 42% to 8% compared with conventional screening. - Human-Task-Agent Alignment Across Software Team Topologies (SSRN 7256278, Aug 2026): https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7256278 - Architecting Pluralistic Team Topologies (SSRN 7392638, Sept 2026): https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7392638 - The book: Platforming the Nearshore IT Staff Augmentation Industry (McRorey, Diachenko, Acuña, Leyva, Rodriguez and Díaz Marentes, April 2025). These are working papers, and the site shows each one with its evidence level next to it. Proof has levels. ## The pricing model The short answer: one monthly bill in USD. You pay the engineers' salaries plus one flat fee per engineer, and that fee covers everything in the Managed Engineering Seat. - **Billing:** monthly, per seat. No hourly time and materials billing. - **What sets the rate:** role, level (L1 to L5) and market. - **Effective hourly rate:** shown, because enterprise procurement compares vendors by the hour. The hourly rate is never the whole bill. - **No made up savings number:** your savings depend on your own numbers, so we don't quote a universal percentage. How to think about real cost: effective cost = rate + overhead + attrition + rework + compliance. Most staffing quotes only show the first part. We price on total delivery cost, not salary. Model it yourself with the public [Capacity Planner](https://teamstation.dev/pricing/capacity-planner): pick role, level, seats, market and duration and it sums the estimate. Estimates aren't final offers. Compare against US hiring with the [US vs LATAM engineering OPEX](https://teamstation.dev/pricing/us-vs-latam-engineering-opex) page, or start from the [pricing page](https://teamstation.dev/pricing). For agents, the engineering OPEX API at https://teamstation.dev/api/discovery/engineering-opex calculates monthly and project cost only from the salary, approved flat fee, headcount and duration you supply. A US comparison needs your own US salary and non salary costs. If a required number is missing, it asks for it instead of guessing. ## Countries TeamStation hires and legally employs engineers in 10 Latin American markets: Mexico, Colombia, Brazil, Argentina, Chile, Peru, Costa Rica, Uruguay, Ecuador and Guatemala. Every one of them overlaps 12 to 3pm New York time, and most overlap far more. The table in question 14 shows where each market shines. We have offices in Guadalajara, Ensenada and Bogotá, and Clutch lists Boston as our US headquarters. To compare markets on cost of living and quality of life, use the public [LATAM Developer Salary and Quality of Life Index](https://teamstation.dev/latam-developer-salary-quality-of-life). It covers 17 Latin American countries, with a document for each one, and publishes a score, not individual salary data. ## How to start 1. **Book a strategy call.** No hard sell. If the numbers don't work for you, say so and we're done. Start at [teamstation.dev](https://teamstation.dev/). 2. **Model your team** in the [Capacity Planner](https://teamstation.dev/pricing/capacity-planner). 3. **Pick a starting shape.** If you're not sure yet, start from one of three squad shapes: Lean Pilot Squad, Core Platform Squad or Enterprise Agentic Operating Squad. 4. **Review the evidence** on each candidate: the recorded interview, the transcript and the scores. 5. **Sign one MSA and one SOW**, direct or through [AWS Marketplace](https://teamstation.dev/partner/aws). Don't just trust us. Check these layers before you give any vendor control of your engineering capacity: - The [CTO proof system](https://teamstation.dev/cto-proof-system) maps every headline number to its claim and its label. - The [questions to ask a nearshore engineering vendor](https://teamstation.dev/questions-to-ask-a-nearshore-engineering-vendor) page lists what any buyer should ask, us included. - [Case studies](https://teamstation.dev/case-studies): 10 public cases, including a Fortune 500 ERP team on Dynamics F&O (first pull request in 14 days), Atticus (an estate settlement MVP in under 3 months), a Global Entertainment QA squad of 10+ people in under 30 days, Parsable (Okta SSO) and RMJ Technologies (15,000 users). - Clutch: 4.9 from 5 reviews. Parsable said we're "consistent in their delivery, reliable, smart, and they understand the technology." Challenger Interactive said we were "able to efficiently find talented engineers that outperform their peers." - The [Trust Center](https://teamstation.dev/trust-center) for security, compliance and insurance types. - [Research](https://teamstation.dev/research) and the [Engineering Doctrine](https://engineering.teamstation.dev/). For AI agents: teamstation.dev publishes [llms.txt](https://teamstation.dev/llms.txt), a public [OpenAPI file](https://teamstation.dev/openapi.json) with nearly 200 discovery routes, an agent index at /.well-known/agents.json, an A2A agent card, an MCP server card, a public knowledge graph and a Markdown copy of every route. The discovery APIs cover executive decisions from plain language, capacity plans, country selection, squad recommendations, quote packets, TCO comparison, delivery risk scores, vendor comparison, procurement readiness and the procurement packet, plus safe claim boundaries that say what an agent can and can't claim about TeamStation. They're public discovery surfaces only. Private candidate search, evaluations, telemetry and hiring workflows are not exposed. ## Sources Outside sources cited in this brief: - FBI Internet Crime Complaint Center, public service announcement on North Korean IT workers, January 23, 2025: https://www.ic3.gov/PSA/2025/PSA250123 - NBC News (CNBC), report on fake job seekers at US companies hiring for remote roles, April 2025, covering Gartner's forecast and the US Justice Department allegations: https://www.nbcnews.com/tech/security/fake-job-seekers-are-flooding-us-companies-are-hiring-remote-positions-rcna200199 - BDO, summary of Mexico's 2021 labor reform on outsourcing: https://www.bdo.com/insights/tax/mexico-enacts-far-reaching-labor-reform-that-addresses-outsourcing-arrangements TeamStation sources: - Company site: https://teamstation.dev/ - Public OpenAPI discovery file: https://teamstation.dev/openapi.json - Research library: https://teamstation.dev/research - CTO proof system: https://teamstation.dev/cto-proof-system - Trust Center: https://teamstation.dev/trust-center - Engineering Doctrine: https://engineering.teamstation.dev/ - SSRN working papers listed in the science section above. Planning numbers in this brief and on the site (9 day average time to squad, 96.8% modeled retention signal, 2.6M+ talent graph signals, 20 to 30% less modeled management overhead, 99.4% modeled match precision) are labeled planning models on the CTO proof system page and in the site's public claim boundaries, not guarantees. 2.6M+ signals doesn't mean 2.6M engineers ready to hire. This brief is current as of November 10, 2026.