Opentag review in 2026: Github, LinkedIn, Company, Ai, Careers, User Experience and FAQs

What Is OpenTag?

If you have been keeping tabs on the AI startup space in 2026, there is a good chance you have come across the name OpenTag. Backed by Y Combinator as part of the Summer 2026 batch, OpenTag markets itself as a model-agnostic AI coworker that lives directly inside Slack and Microsoft Teams. The idea is straightforward: instead of switching between a dozen different tools, you tag OpenTag in a channel thread, hand it a task, and it delivers the result right there for the whole team to see.

The company was founded by three UC Berkeley alumni. Tony Kam, who serves as CEO, previously worked at Tesla and Intel before co-founding Lilac Labs, which was part of YC's S24 cohort. Shelden Shi, the CTO, brings experience from Flatiron Health and Roche. Wilson Nguyen, the Chief Engineer, came from Hewlett Packard Enterprise. All three worked together at Lilac Labs before pivoting into what would become OpenTag.

On paper, this sounds promising. A small, focused team with serious pedigree, tackling a real pain point in enterprise collaboration. But as we always do at ICON POLLS, we wanted to look beyond the pitch deck and the YC badge. We spent time examining their product, their online presence, their career offerings, their GitHub activity, and what actual users are saying about the experience. Here is what we found.

OpenTag Company Overview

OpenTag operates under a fairly lean structure. As of mid-2026, the team consists of just three people, all of whom are co-founders. The company is headquartered in San Francisco and positions itself squarely within the enterprise AI and workflow automation space.

Their core product is an AI agent that was originally launched under the working name Gini. It integrates into team chat environments and is designed to go beyond simple question-and-answer interactions. According to their marketing materials and Y Combinator profile, OpenTag can take on multi-step tasks, connect to third-party tools like HubSpot, Gmail, and Stripe, and even join calls. Every task runs on a sandboxed machine in the cloud, and the company emphasizes that sensitive actions require human approval through built-in gates.

One of their more interesting selling points is the automatic wiki feature. As OpenTag handles tasks and observes how a company operates, it gradually builds a knowledge base of how decisions get made, who owns what, and which workflows are repeated. The goal is to create a living, self-updating company brain that stays current without anyone having to manually maintain it.

The business model is credit-based. Teams buy monthly credits, and each job or task consumes a portion of those credits. This is fairly standard for AI tool pricing in 2026, though the exact cost per task was not clearly published on their website at the time of our review.

OpenTag AI: How It Actually Works

The biggest technical differentiator OpenTag pushes is its model-agnostic architecture. Unlike Anthropic's Claude Tag, which is locked to Claude models, OpenTag claims to route tasks to whichever AI model is best suited for the job. Their website mentions access to roughly 80 different models, and they advertise potential cost savings of up to 70% compared to single-model competitors.

This approach has its advantages. When a better model launches from any provider, OpenTag users theoretically benefit from those improvements automatically. It also means the platform is not reliant on a single AI vendor's pricing structure or performance limitations.

However, we noticed a few concerns. First, there is limited public documentation about how the model routing actually works. Which models are used for which tasks? How does the system decide? These are questions that enterprise buyers will inevitably ask, and the answers were not easy to find during our research. Second, being model-agnostic introduces complexity around consistency. If your AI coworker switches between different models across different tasks, the quality and tone of responses could vary in ways that feel unpredictable to end users.

On the Product Hunt launch page, user feedback was generally positive but thin. A few early adopters praised the auto-wiki feature and the idea of a Slack-native AI coworker, but there were very few detailed case studies or performance benchmarks to reference. For a product this early, that is somewhat expected, but it still makes it difficult to evaluate real-world reliability.

OpenTag on GitHub

When we looked into OpenTag's GitHub presence, things got a bit complicated. There is no single, prominent public repository tied directly to the tryopentag.com product. The company appears to keep its core product closed-source, which is understandable for a venture-backed SaaS startup.

That said, the broader OpenTag name has significant presence on GitHub, though most of it belongs to unrelated projects. There is an open-source organization called OpenTagOS that maintains several repositories focused on PHP tooling, Terraform providers, and testing frameworks. Separately, multiple open-source projects have adopted the OpenTag name as a nod to the agent-mention workflow that Claude Tag popularized. These include forks by CopilotKit and independent developers who have built open-source alternatives for routing coding agents through Slack and GitHub threads.

For anyone searching for OpenTag on GitHub, it is important to distinguish between the YC-backed commercial product at tryopentag.com and these community projects. They share a name and a general philosophy, but they are not the same thing. The commercial OpenTag does not appear to have a public codebase, open-source SDK, or developer documentation hosted on GitHub at this time.

From a transparency perspective, this is a gap. Developers and technical evaluators often look at GitHub activity as a signal of how seriously a company invests in its engineering community. The absence of any public repository tied to the actual product may raise eyebrows among more technical buyers.

OpenTag on LinkedIn

OpenTag maintains a LinkedIn company page under the handle tryopentag. At the time of our review, the page was relatively new and had a modest following, which makes sense given that the company just launched in 2026 with a three-person team.

The LinkedIn page displays standard company information and links back to the main website. Posts appear to focus on product announcements, the Y Combinator demo day, and early launch milestones. There is not a lot of thought leadership content or in-depth technical writing being shared yet, which is something we would expect to see develop as the company grows.

It is worth noting that there are other companies using the OpenTag name on LinkedIn as well. One is a technology services firm based in Malta that was founded in 2014. Another is the iGaming platform provider based in Sofia, Bulgaria, which has a much larger team and a more established LinkedIn presence with hundreds of followers. If you are looking for the AI coworker startup, make sure you are following the right page. Search for tryopentag specifically to avoid confusion.

OpenTag Careers and Hiring in 2026

As of our review, OpenTag has not posted public job openings tied to the AI coworker product. With only three co-founders on board, the company is still in its earliest phase. The Y Combinator profile and Dealroom data both list the team size at three, and the estimated enterprise value sits in the $500K to $750K range, which is typical for a very early-stage startup.

If you are a job seeker looking specifically at OpenTag careers, the timing may be a bit early. Most YC-backed companies begin scaling their hiring after completing the accelerator program and securing follow-on funding. Given that the founders all have strong engineering backgrounds and came from notable companies like Tesla, Intel, Roche, and Hewlett Packard Enterprise, it is reasonable to expect that early hires will lean heavily toward engineering and product development roles.

For anyone interested in joining the team, reaching out directly through LinkedIn or the contact information on the YC profile (tony@tryopentag.com) is probably the most effective route. Do not confuse job listings from the Sofia-based iGaming OpenTag with this company. Those are entirely separate organizations.

User Experience

The user experience side of OpenTag is where things feel both promising and underdeveloped at the same time. On one hand, the core concept is strong. Tagging an AI assistant directly inside Slack or Teams reduces friction significantly. There is no separate dashboard to log into, no context switching, and results show up in the exact thread where the conversation is happening. That is genuinely useful.

The auto-wiki feature is another highlight. The idea that your AI coworker passively builds and updates a company knowledge base just by observing how work gets done is compelling. It solves a real problem that every growing team faces: documentation that goes stale the moment someone writes it.

On the other hand, we noticed some things that need attention. The website at tryopentag.com, while clean, does not offer much in the way of detailed product documentation, pricing breakdowns, or tutorial content. For a tool that is asking teams to trust it with real workflows and third-party tool access, the onboarding materials feel thin. There is a demo booking option, which suggests the company is still operating in a high-touch sales model rather than a self-serve experience.

We also could not find a free trial or sandbox environment where potential users could test the product before committing. The Product Hunt launch mentioned a 50% discount for two months to celebrate the launch, but beyond that, pricing details remain vague. Credit-based billing can be tricky if users do not have a clear picture of how many credits each task type consumes.

Overall, the user experience has a solid foundation, but it feels like a product that is still being shaped rather than one that is ready for broad adoption. Early adopters who are comfortable with that level of ambiguity will probably get the most out of it right now.

What We Liked

The model-agnostic approach is a smart differentiator. Not being locked into a single AI provider gives teams flexibility and potentially lower costs as the market evolves. The Slack and Teams integration keeps everything in one place, which is exactly where modern teams already spend their time. The founding team brings real credibility with backgrounds from Tesla, Intel, Roche, and HP Enterprise, plus prior startup experience through Lilac Labs and two rounds of Y Combinator participation. The automatic wiki concept is genuinely innovative and addresses a very common pain point around institutional knowledge management.

What Gave Us Pause

The company is extremely early-stage with just three employees and limited public traction. There is very little public documentation, no developer SDK or open-source presence tied to the actual product, and pricing transparency is lacking. The user base appears very small, with only about 10 teams in early access at the time of the YC launch. There are multiple unrelated companies using the OpenTag name, which creates brand confusion that could hurt discoverability and trust. The Glassdoor reviews that show up under the OpenTag name belong to the iGaming company in Sofia, not this startup, which can mislead job seekers and researchers alike.

Frequently Asked Questions (FAQs)

1. What is OpenTag and what does it do?

OpenTag is an AI-powered coworker that integrates directly into Slack and Microsoft Teams. It is designed to handle tasks, answer questions, join calls, and automate repetitive workflows without requiring users to leave their team communication platform. The company was founded in 2026 and is backed by Y Combinator.

2. Who founded OpenTag?

OpenTag was co-founded by Tony Kam (CEO), Shelden Shi (CTO), and Wilson Nguyen (Chief Engineer). All three are UC Berkeley alumni who previously worked together at Lilac Labs, which was also part of the Y Combinator S24 batch. Tony Kam has experience from Tesla and Intel, Shelden Shi worked at Flatiron Health and Roche, and Wilson Nguyen came from Hewlett Packard Enterprise.

3. Is OpenTag the same as Claude Tag?

No. Claude Tag is Anthropic's proprietary AI teammate for Slack, and it only works with Claude models. OpenTag positions itself as a model-agnostic alternative, meaning it can route tasks to multiple AI models from different providers. The two products share a similar user interface concept but are built by separate companies with different approaches.

4. How much does OpenTag cost?

OpenTag uses a credit-based billing model where tasks consume credits from a monthly balance. As of our review in September 2026, detailed pricing was not publicly listed on their website. They offered a 50% discount for the first two months as a launch promotion. Interested teams can book a demo through the website for specific pricing details.

5. Is OpenTag open source?

The commercial product at tryopentag.com is not open source. However, several unrelated open-source projects on GitHub use the OpenTag name and offer similar agent-mention workflows. These community projects are not affiliated with the YC-backed OpenTag company.

6. Does OpenTag have a GitHub repository?

The commercial OpenTag product does not appear to maintain a public GitHub repository or open-source SDK at this time. The GitHub accounts and repositories you may find under variations of the OpenTag name belong to separate, unrelated projects and developers.

7. Is OpenTag hiring in 2026?

As of mid-2026, OpenTag has a three-person team consisting entirely of its co-founders. There are no publicly listed job openings tied to the AI coworker product. Hiring is expected to begin as the company scales after completing the Y Combinator program. Interested candidates should reach out through LinkedIn or the founders' contact information listed on the YC profile.

8. Is OpenTag safe to use for sensitive business data?

OpenTag states that each task runs on a sandboxed cloud machine that gets torn down after completion. The platform also includes approval gates for sensitive actions, requiring human sign-off before certain operations are executed. However, since the company is very early-stage and detailed security documentation is not publicly available, teams handling highly sensitive data should request a thorough security review before deployment.

9. What tools does OpenTag integrate with?

According to the company's product materials, OpenTag connects with third-party tools including HubSpot, Gmail, and Stripe, among others. The integrations are handled through sandboxed cloud machines that connect to these services during task execution. The full list of supported integrations was not publicly documented at the time of this review.

10. How does ICON POLLS rate OpenTag overall?

ICON POLLS gives OpenTag an overall rating of 3.0 out of 5.0, placing it in the average range. The product vision and founding team are strong, but the extremely early stage of development, limited user base, lack of public documentation, and brand confusion with other companies of the same name all contributed to a moderate score. It is a company to keep an eye on, but it is not yet ready for a strong endorsement.