{"@context":"https://schema.org","@type":"JobPosting","title":"Technical Co-Founder with Investment","description":"

Softworker AI is building a multiplayer collaborative workspace where human teams and autonomous AI agents execute business work together, with shared context, permissions, approvals and governance.

Softworker AI

is building a multiplayer collaborative workspace where human teams and autonomous AI agents execute business work together, with shared context, permissions, approvals and governance.

Product walkthrough: https://www.youtube.com/watch?v=Gzo7NnqYHsg

I am a technical founder with 15 years of engineering experience, and the MVP development is in progress. But the road ahead is a set of open research problems that need to become production infrastructure, and that is more than one person can do well.

Frontier AI labs have taken coding to a level where agents are productive collaborators. We want to do the same for every other business process, with a new harness engine where multiplayer is native rather than bolted on, built for strict enterprise environments that are sometimes air-gapped and always bound by data-residency requirements.

Here is what we need to build:

1. An in-house agent runtime. Existing frameworks do not meet our governance requirements for permissions, approvals and auditability. We need our own harness engine, designed from day one for on-premise and air-gapped deployment.

1. An in-house agent runtime.

Existing frameworks do not meet our governance requirements for permissions, approvals and auditability. We need our own harness engine, designed from day one for on-premise and air-gapped deployment.

2. A multiplayer chat product. Humans and agents are first-class participants in the same conversation. The same rules govern who can see what, who can act, and who must approve.

2. A multiplayer chat product.

Humans and agents are first-class participants in the same conversation. The same rules govern who can see what, who can act, and who must approve.

3. Multi-agent shared memory with per-participant context. Research still lists memory consistency, leakage, stale propagation and provenance collapse as unsolved failure modes of shared agent memory. We need to allocate context to every participant, human or agent, so no single actor's view becomes skewed, and make it model-independent so it stays relevant regardless of which models win.

3. Multi-agent shared memory with per-participant context.

Research still lists memory consistency, leakage, stale propagation and provenance collapse as unsolved failure modes of shared agent memory. We need to allocate context to every participant, human or agent, so no single actor's view becomes skewed, and make it model-independent so it stays relevant regardless of which models win.

4. Agent identity and security runtime. Static identity does not work for agents: they are ephemeral, spawned and destroyed constantly, and act on behalf of humans through delegation chains that standards like OAuth and MCP are still catching up to. We need a runtime that keeps dynamic, delegated identity coherent across sessions, restarts and chains of agents, with no internet connection.

4. Agent identity and security runtime.

Static identity does not work for agents: they are ephemeral, spawned and destroyed constantly, and act on behalf of humans through delegation chains that standards like OAuth and MCP are still catching up to. We need a runtime that keeps dynamic, delegated identity coherent across sessions, restarts and chains of agents, with no internet connection.

5. Support for third-party agents. Enterprises will run agents from many vendors, and protocols like A2A now let them collaborate across frameworks, but that opens governance gaps. Our runtime must let an external agent join a shared conversation as a governed participant: identified, scoped, approved and audited like everyone else.

5. Support for third-party agents.

Enterprises will run agents from many vendors, and protocols like A2A now let them collaborate across frameworks, but that opens governance gaps. Our runtime must let an external agent join a shared conversation as a governed participant: identified, scoped, approved and audited like everyone else.

6. An evaluation framework for shared conversations. Most evaluation scores a single model's single output, and current benchmarks under-measure governance failures like delegation drift. We need to track the quality of the whole shared conversation over time, humans and agents together, including where governance broke.

6. An evaluation framework for shared conversations.

Most evaluation scores a single model's single output, and current benchmarks under-measure governance failures like delegation drift. We need to track the quality of the whole shared conversation over time, humans and agents together, including where governance broke.

Each of these could be a company on its own. We are building them as one coherent system because that is what the enterprise use case demands.

Who We're Looking For

We are looking for a Technical Co-Founder who will bring both strong capital and relevant expertise, and help turn SoftworkerAI into a category-defining company. This is an equity-based co-founder opportunity, not a salaried job.

This is an equity-based co-founder opportunity, not a salaried job.

You should:

Be based in the UAE(or willing to relocate) and have(can arrenge) needed work permit and residence VISAHave deep, hands-on engineering expertise in one or more of: distributed systems, agent runtimes, identity and access management, memory and retrieval infrastructure, or enterprise/security softwareBe able to invest meaningful capital into the company alongside your timeBe building or have previously built an AI company or productBe ready to join full-time, not alongside a job, degree or another startupWant to combine forces with a complementary founderBe ambitious, resourceful, AI-literate and comfortable starting from zero

Be based in the UAE(or willing to relocate) and have(can arrenge) needed work permit and residence VISA

Have deep, hands-on engineering expertise in one or more of: distributed systems, agent runtimes, identity and access management, memory and retrieval infrastructure, or enterprise/security software

Be able to invest meaningful capital into the company alongside your time

Be building or have previously built an AI company or product

Be ready to join full-time, not alongside a job, degree or another startup

Want to combine forces with a complementary founder

Be ambitious, resourceful, AI-literate and comfortable starting from zero

What You'll Own

Co-owning technical architecture and engineering execution across the runtime, memory, identity, interoperability and evaluation layersShipping the MVP and iterating with design partnersEnterprise deployment: air-gapped installs, data residency, security postureHiring and leading the early engineering teamTechnical due diligence and investor conversations alongside the founding team

Co-owning technical architecture and engineering execution across the runtime, memory, identity, interoperability and evaluation layers

Shipping the MVP and iterating with design partners

Enterprise deployment: air-gapped installs, data residency, security posture

Hiring and leading the early engineering team

Technical due diligence and investor conversations alongside the founding team

Get to Know Process:

Round 1: Application. Send your CV, LinkedIn, references and details of what you have built, including your product, demo or repository.

Round 1: Application.

Send your CV, LinkedIn, references and details of what you have built, including your product, demo or repository.

Round 2: Written Take-Home Assignment. A long, essay-style assessment on values, ambition, decision-making, conflict, resilience, financial expectations and founder readiness; AI use allowed if disclosed.

Round 2: Written Take-Home Assignment.

A long, essay-style assessment on values, ambition, decision-making, conflict, resilience, financial expectations and founder readiness; AI use allowed if disclosed.

Round 3: Technical Assignment. Pick one or two of the six problems above and write a concise review of how you would approach them for air-gapped enterprise deployment and prioritise the next six months.

Round 3: Technical Assignment.

Pick one or two of the six problems above and write a concise review of how you would approach them for air-gapped enterprise deployment and prioritise the next six months.

Round 4: Online Technical and Culture-Fit Interview. Walk us through your technical review live, and stand behind your Round 2 answers on working style, communication and disagreement.

Round 4: Online Technical and Culture-Fit Interview.

Walk us through your technical review live, and stand behind your Round 2 answers on working style, communication and disagreement.

Round 5: Founder-Alignment Conversation. A transparent talk on vision, roles, full-time commitment, capital contribution, equity, decision-making, personal and financial circumstances and long-term expectations.

Round 5: Founder-Alignment Conversation.

A transparent talk on vision, roles, full-time commitment, capital contribution, equity, decision-making, personal and financial circumstances and long-term expectations.

Round 6: Two Weeks Together in Dubai. Work with us in person on real priorities such as shipping the MVP, runtime design, design-partner onboarding and fundraising, after which all sides decide whether to build the company together.

Round 6: Two Weeks Together in Dubai.

Work with us in person on real priorities such as shipping the MVP, runtime design, design-partner onboarding and fundraising, after which all sides decide whether to build the company together.

How to Apply

Email career@softworker.ai with:

Your CV and LinkedIn profileYour AI product, demo, website or repository

Your CV and LinkedIn profile

Your AI product, demo, website or repository

This is not a conventional job application. It is our attempt to find a co-founder carefully, honestly and with enough real-world evidence to avoid relying on vibes alone.

Note: Applying through LinkedIn alone will not be considered a complete application.

","url":"https://jobhunt.work/app/jobs/4470564571?provider=1","identifier":{"@type":"PropertyValue","name":"Softworker AI","value":"4470564571"},"hiringOrganization":{"@type":"Organization","name":"Softworker AI","logo":"https://www.linkedin.com/company/softworker-ai/life"},"datePosted":"2026-09-26T16:09:09.000Z","jobLocation":{"@type":"Place","address":{"@type":"PostalAddress","addressCountry":"AE"}},"sameAs":"https://www.linkedin.com/jobs/view/4470564571"}