Company status
1AYM is an OpenAI Select Partner
1AYM is an OpenAI Select Partner. Company status and personal certifications are different things: Select Partner is a company status awarded after a partner agreement, a compliance review and a technical assessment, not a certification from OpenAI and not a partner claim with Anthropic; the two Claude credentials are personal, held by Tayyeb Mahmud, founder: Claude Certified Architect – Professional (23 July 2026) and Claude Certified Associate – Foundations (21 July 2026), both verifiable on Credly. We are model-selective in production across OpenAI, Anthropic, Gemini and open-source. Client environments run in daily production, for clients across the UK, the US and the Gulf.
The status
What is an OpenAI Select Partner
A company status in the OpenAI Partner Network, held by 1AYM rather than by any individual.
OpenAI Select Partner is a company status, awarded after a partner agreement, a compliance review and a technical assessment. It belongs to 1AYM. It is written that way on every surface of this site.
A partner status is not a certification, and it is not a sales agreement that decides which model we put in production. Model choice still comes from evidence on the workload in front of us.
The work
What an OpenAI implementation partner should deliver.
A badge is not a system. The test is whether a real workflow is running in your environment, governed, measured and handed over.
Architecture
Target architecture, model and tool routing, access control and the operating model, written down so the reasoning outlives the engagement.
One real workflow
End to end, in your systems and against your data. A demonstration that skips the hard integration proves nothing.
Governance
Least-privilege access, audit logging, deterministic checks and human approval sized to the risk, built as controls rather than written as a policy.
Evaluation in CI
Regression suites that run when a prompt or a model changes, so a quality drop is caught before users find it.
Enablement
The people who will run it are in the work from the start, so the capability stays when the engagement ends.
Handover
Documentation, training and a written handover. The measure is what your team can maintain, change and build on once we have gone.
The same six points as evidence. At handover, each deliverable should be something you can open, run, or ask someone on your own team about.
| Dimension | What you should be able to see | What should worry you |
|---|---|---|
| Architecture (written) | A document naming the models, the tool routing, access and the operating model, with the reason and the date for each choice. | Decisions that live in a slide deck or in one engineer's head. |
| One real workflow | The workflow live in your environment, on your data, used by the people it was built for. | A demonstration in the partner's own account, or a pilot that never touched your real systems. |
| Governance: access and audit controls | Least-privilege roles, audit logs your team can query, and human approval on actions that carry risk. | A usage policy with no control behind it. |
| Evaluation in CI | A regression suite that runs when a prompt or a model changes, with results your team can read. | Quality checked once, by hand, before launch. |
| Enablement: trained owners | Named people on your side who have already changed the system, with the partner beside them. | Training delivered as one session in the final week. |
| Handover | Documentation, a runbook and a written handover, with the partner's access narrowed or removed. | Anything that depends on a system only the partner can reach. |
The same three stages as the rest of 1AYM's work, in this order. Each one ends with something written down and handed over.
Discover
Typically two to four weeks with your sponsor and the people who do the job today. We agree what the outcome is worth, what stands in the way, and how it will be judged.
You leave with a ranked list of what to build, the architecture for it, and an investment case with real numbers in it.Ship a working slice
One real workflow, end to end, in your systems and against your data. A demonstration that skips the hard integration proves nothing, so we do not build one.
You leave with working software in your environment, and evidence of what it does.Run it, governed
Access rules, approval gates, monitoring and training, then a staged rollout. Your team runs it. We measure adoption, cycle time and cost, so the change is reported rather than asserted.
You leave with a running system, a team that can maintain it, and numbers for the board.
Proof
Two environments, in daily production.
Every figure below is already published on this site, with its source. Nothing is added here.
A government-accredited EdTech in the Middle East
A whole-interview speaking assessment on OpenAI audio models with a Gemini fallback, on an estate whose database was replatformed onto Postgres in Google Cloud's Doha region (me-central1) to meet Gulf data-residency requirements.
Engagement file D-02
A large international marketing agency
We lead the architecture of the enterprise AI platform underneath the agency's organisation-wide AI programme. That platform is featured in a published Anthropic customer case study: about 1,000 employees reached, around 400 employee-authored skills in the first four weeks, 91% adoption in North America.
Engagement file D-01
Model choice
Model-selective, in production.
The choice comes from what we run across OpenAI, Anthropic, Gemini and open-source, not from a vendor deal. We are model-selective: the model is picked per workload rather than by loyalty.
In production today: OpenAI audio models for speech assessment with a Gemini fallback, Claude for enterprise platform and agent estates, Codex and Claude Code for engineering work. Where a model is a poor fit we will say so.
The same applies to the tools we compare. Our comparison of Codex and GitHub Copilot, for example, works from each vendor's own documentation and matches a tool to a team's situation rather than naming a winner.
Data and law
UK GDPR and the Data Protection Act 2018.
We work to UK GDPR and the Data Protection Act 2018, plus whatever your sector adds, and we architect for data residency where a region is a requirement.
We hold no ISO or SOC certification, and we do not imply one.
The Gulf
OpenAI in the Gulf.
For a Gulf buyer, the first OpenAI question is usually where the data goes. On OpenAI's published terms, the UAE is a data-residency region for ChatGPT Enterprise and for the API, including in-region processing, with approvals and limits. Saudi Arabia, Qatar and the other Gulf states are not on OpenAI's region list, so for buyers there residency is a design question.
In ChatGPT, OpenAI's pricing page lists the UAE among ten data-residency regions [2], and its help centre lists it among the three regions, with Europe and the US, where ChatGPT Enterprise inference can also stay in-region. Sign-in, routing and anything sent through an external integration may still be processed outside the region [3].
On the API, the UAE region covers storage and processing, but only for some models, and only once OpenAI has approved the account for it [4]. The newest model a team wants may not be one the region serves.
Our part is the work around those terms. Where data rests and where inference runs are different facts, so a sound design names the region of every data store and every model call separately, and lets the region decide the model: if the model you want is not served there, you should hear it before the build starts. We architect for residency where a region is a requirement: one production estate we run was replatformed onto Postgres 16 in Google Cloud's Doha region (me-central1) to meet Gulf data-residency requirements. 1AYM has also built bilingual Arabic-English search and Arabic document OCR, has an entity in the UAE and works with clients across the Gulf. Our guide to AI data residency in the Gulf sets out what each country's law asks for.
This is a reading of OpenAI's published terms on the dates shown, not legal advice. Which rule applies to which data set is a question for your counsel, and vendor terms change, so check them again before a contract depends on them.
For engineers: the UAE region on the OpenAI API
Data residency on the API is set per project: pick the region when the project is created, then send its requests to the regional domain, ae.api.openai.com for the UAE. A single request can also ask for regional processing through the prefixed domain, using a key from a project in the Global geography, where the endpoint and the model support it [4].
In the UAE, OpenAI lists in-region processing on the Responses and Chat Completions endpoints for GPT-5.6 Luna, GPT-5.5 (the 23 April 2026 snapshot), GPT-5.5 Pro (the 23 April 2026 snapshot, Responses only) and GPT-5.2 (the 11 December 2025 snapshot), and for text-embedding-3-large. None of the GPT-6 models is on that list. Processing for the web search, file search and Code Interpreter tools is listed for the US and Europe only [4].
Any region other than the US needs approval for abuse monitoring controls and a Modified Retention amendment, and the UAE needs a further approval on top. Eligible models released on or after 5 March 2026 carry a 10% uplift on residency endpoints. Residency covers customer content: system data such as account details, usage and billing may be processed and stored outside the region [4].
Log the region and the model on every call, so a residency claim can be shown from the records rather than remembered.
How you buy
Five named engagements. No OpenAI-only SKU.
Implementation work is bought as one of the five engagements already on this site. There is no separate OpenAI product line, and no seat package. The status does not change what you can buy.
For C-suite and transformation leads
AI Opportunity & Feasibility Sprint
Find out which two or three workflows are worth building, what each is worth and what it would cost, before you commit a budget to any of them.
Executive AI discovery
For CIO, CTO, CDO, operations and finance
Enterprise AI Platform & Agentic Workflows
Turn scattered pilots into one governed place for AI to run, then put real workflows on it. Access rules, checks and human approval are built in, so the second workflow does not start from scratch the way the first one did.
Production AI systems
For engineering leaders
AI Engineering Transformation
Your engineers are already using coding agents such as Codex and Claude Code, with or without a policy. This puts review, guardrails and CI around that, so the code it writes reaches production the same way every other change does.
AI harness & platform engineering
For scale-ups and enterprises
Fractional AI Implementation Team
Senior architecture, hands-on implementation and enablement without hiring a permanent AI platform team. We own the technical decisions, build with your engineers and leave them able to run it.
Fractional AI team
For programmes that need senior capacity now
Embedded Engineers on Contract
A senior engineer, or a team, inside your programme on contract. You direct the work day to day, and we hold the standard every 1AYM engineer is hired against, plus the cover behind them.
Embedded engineers
If your company already holds ChatGPT Enterprise seats, getting them used and signed off by security is setup and enablement work, not another purchase. Our guide to the first 90 days of rolling out ChatGPT Enterprise sets out the order, from sign-in to the day-90 review.
Public-sector buyers can work with 1AYM through G-Cloud via our partner BAPRO. BAPRO is the delivery partner and 1AYM is the AI partner, across everything 1AYM offers. Our page on AI for the UK public sector explains how buying through G-Cloud works.
The people
Who does the work.
Every 1AYM engineer is either certified on the platforms we build on, or has shipped inside a top-tier engineering organisation. The people who write the strategy write the code, and we scale the team to the contract.
The senior engineers a client meets first: a Principal Architect (10+ years; AI adoption, agent systems, Claude and OpenAI platform architecture; holds both Anthropic Claude certifications; in production at the client environments above); a Principal Engineer (14 years, London; data platform, context and memory architecture); a Principal Data Engineer (17 years, London; pipeline orchestration, CI/CD, data infrastructure).
1AYM delivers the work rather than any one person. Colleagues are named only with their permission, and you meet the engineers before you sign.
Meta, Spotify, Viasat/Inmarsat, Starling Bank, UBS, S&P Global, Sky and Société Générale are former employers of our engineers, not clients of 1AYM.
Before you sign
A buyer checklist.
Questions that separate an implementation partner from a seat reseller. Each one is answerable from this page or the file it links.
- 01
Can they show a system your people would depend on every working day, not a demonstration and not unused seats?
- 02
Is model choice a workload decision, with more than one provider already in production?
- 03
Will the first delivery be one real workflow in your systems, with architecture written down beside it?
- 04
Are access, audit, deterministic checks and human approval built as controls, and does evaluation run in CI?
- 05
Is enablement inside the engagement, with a written handover, so your team can run it afterwards?
- 06
Is the commercial shape one of the five named engagements, rather than an OpenAI-only package?
Questions
Before you book.
What does Select Partner status actually mean?
It is a company status in the OpenAI Partner Network, held by 1AYM rather than by any individual, and awarded after a partner agreement, a compliance review and a technical assessment. It is not a certification, and it does not decide which model we put in production.
Is a partner status the same as a certification?
No. A partner status is awarded to a company. A certification is issued to a person who has sat a proctored exam. We keep them apart, in two sentences rather than one phrase: 1AYM is an OpenAI Select Partner. Our founder and principal architect holds Anthropic's Claude Certified Architect credential.
Which models do you run in production?
Whichever the workload argues for, across OpenAI, Anthropic, Gemini and open-source. In production today: OpenAI audio models for speech assessment with a Gemini fallback, Claude for enterprise platform and agent estates, Codex and Claude Code for engineering work.
How do we buy this? Is there an OpenAI-only package?
No. Implementation work is bought as one of the five named engagements already on the site: a feasibility sprint, an architecture and production build, an engineering enablement programme, a fractional AI implementation team, or embedded engineers. There is no separate OpenAI product line.
Is an OpenAI partner the same as a reseller?
No. A reseller sells licences; an implementation partner builds on the licences you hold and hands the result over. 1AYM does not resell OpenAI licences. Your ChatGPT or API agreement stays between your organisation and OpenAI, and the workspace, the admin console and the bill stay yours. Our engineers work under identities your organisation issues, so their access is yours to grant and to remove.
Who can help deploy ChatGPT Work in a company?
Three parties share it. OpenAI sells the plan and holds the contract with you. Your workspace owners and admins decide who gets Work, by role, with separate settings for cloud tasks and for local work in the desktop app; on eligible Enterprise workspaces it is on by default for members on the default role unless an admin turns it off (OpenAI Help Center, checked 29 September 2026). An implementation partner does the rollout: roles and connectors decided before switch-on, one real workflow per team, usage tracked and people trained. Setting up and rolling out AI work tools, ChatGPT Work among them, is core work for 1AYM.
Do you hold ISO or SOC certification?
No. We work to UK GDPR and the Data Protection Act 2018, plus whatever your sector adds, and we architect for data residency where a region is a requirement. We hold no ISO or SOC certification, and we do not imply one.
Sources
- [1]OpenAI Help Center, ChatGPT Work and Codex: workspace access to Work, by role, cloud and local (checked 29 September 2026)
- [2]OpenAI, ChatGPT pricing: security features on business plans, including the data-residency regions (checked 29 September 2026)
- [3]OpenAI Help Center, Data residency and inference residency for ChatGPT (checked 29 September 2026)
- [4]OpenAI API documentation, Data controls in the OpenAI platform: data residency, regions, eligible models and conditions (read 30 September 2026)
Bring one workflow.
Half an hour is enough to tell you whether we can help, what it would take, and whether a production system is the right buy. We will say plainly when the answer is that you do not need us.
Last reviewed · 1AYM