Back to News
ProductAI Understanding briefing

Wrike founder launches Hops, an AI‑driven workspace that forces engineers to code at speed

Andrew Filev’s new startup Hops uses AI agents to generate pull requests and automate team workflows, promising ten‑fold productivity gains and a public launch in December.

4 min readRead the original reporting
Source-provided image accompanying Wrike founder launches Hops, an AI‑driven workspace that forces engineers to code at speed
Attributed reportingSource recorded
Publisher
fortune.com
Source type
Reporting by a news outlet — not a first-party document.

What we could not confirm independently: This claim is attributed to the named outlet. We did not verify it against a first-party document. (fortune.com)

ContextUnderstand this in 60 seconds

Key terms

Generative AI
AI systems that produce new content such as text, images, audio, video, or code.
Embedding
A numeric vector representation that captures semantic meaning of text, images, or other data.
Feature
An input variable used by a model to make predictions.

What happened

Filev’s startup Hops, billed as an “AI operating system for your business,” opened to the public on Dec. 1 after an early‑access trial with more than 2,500 teams. The platform embeds AI agents that automatically generate code changes (pull requests) and perform tasks across chat, docs, spreadsheets and integrations with roughly 2,000 external services. Engineers must review each AI‑generated PR the same day, a rule that pushed the team’s merged PR volume from 798 in Q1 to 4,211 in Q3—a 5.3× increase. Filev says the speed boost translates to at least a 10× productivity gain for developers and has already helped his company cut $135,000 in SaaS subscriptions. Hops also includes a “meeting‑agent” beta that can answer questions and generate slide content, slated for general availability in December.

Hops launched publicly on Dec. 1 following an early‑access program that involved over 2,500 teams testing the platform’s AI‑driven features. The product combines typical collaboration tools—chat, shared documents, spreadsheet‑style tables, and file storage—with AI agents that can act within those environments.

The AI agents automatically generate code changes, creating pull requests (PRs) that engineers must review the same day. This rule forced the engineering team to accelerate its workflow, resulting in a rise from 798 merged PRs in Q1 to 4,211 in Q3, a 5.3× increase. Adjusted for code size, the increase is 6.3×. Filev estimates a ten‑fold speedup in overall developer productivity, on top of a prior 2× gain from earlier coding agents.

Beyond code, Hops integrates with roughly 2,000 external services (e.g., Google, Meta ad accounts, Stripe). Users can query the system for analytics—such as weekly ad‑spend insights—without manually pulling data. The platform also lets teams build custom internal apps, enabling the company to cancel subscriptions to tools like Tableau, Jira, Notion, Slack and HubSpot, saving about $135,000 annually.

A “meeting‑agent” beta, demonstrated during the interview, can answer questions and generate slide content in real time. Filev says this will be generally available in December, expanding Hops’ AI capabilities from code to meeting assistance.

Source details: fortune.com

Why it matters

Hops illustrates a concrete shift from using as a personal assistant to it directly into the fabric of team software. By automating code reviews, data‑pulls and routine SaaS tasks, the platform promises to close the gap identified by MIT’s NANDA study, which found 95 % of organizations getting no return on generative‑AI spend. If the claimed productivity gains hold, enterprises could see faster product cycles, reduced reliance on multiple niche tools, and measurable cost savings on software licences. The model also raises questions about workforce roles: managers are being re‑skilled to oversee AI agents rather than people, and hiring now emphasizes “agency” and comfort with AI‑augmented work. The public rollout gives other firms a real‑world test case for scaling AI‑generated code and workflow automation at enterprise scale.

AI agents directly into collaborative workflows addresses a key pain point highlighted by recent research: organizations often see individual productivity gains from AI but fail to translate those gains into company‑wide results. Hops’ shared‑knowledge model means that improvements made by one employee become instantly available to the whole team.

If Hops delivers on its productivity claims, enterprises could see faster development cycles, reduced time‑to‑market, and lower software‑licensing costs. The platform’s ability to replace multiple SaaS tools with a single AI‑augmented workspace could reshape budgeting and vendor strategies.

The shift in managerial responsibilities—from people‑centric to AI‑agent oversight—signals a broader re‑definition of engineering leadership. Hiring criteria now prioritize “agency” and comfort with AI tools, potentially reshaping talent pipelines in tech.

The public launch provides a real‑world test case for scaling AI‑generated code at enterprise scale, an area that has largely remained experimental. Success or failure will inform how other firms approach AI‑driven development pipelines.

Interactive Mechanism

Interactive Mechanism: How It Actually Works

Explore the underlying technology behind this development interactively.

Agent Lifecycle Stage:
1
User Intent & Planning: "Audit customer refund request #4092 and settle payment."
2
Tool Calling: Emits structured JSON call crm_get_transaction(id='4092').
3
Guardrail & Verification:Paused: High-value action requires human operator sign-off.
4
Final Settlement: Refund recorded, email receipt dispatched, and audit log stored.
Core takeaway: An AI agent is not just a language model—it is a closed loop of planning, tool invocation, and environment feedback. Production systems require self-healing retries and strict human approval guardrails.
Interactive Concept Check+10 Points
AI Models Explained Quiz

Which component of an AI application is the machine-learning model itself?

What to watch next

Key indicators will include adoption rates after the December GA, the stability and quality of AI‑generated pull requests, and how quickly other enterprises adopt similar AI‑agent‑centric workflows. Watch for feedback on the meeting‑agent beta, especially around accuracy and privacy, and for any shifts in hiring practices at Hops that could signal broader industry trends in AI‑augmented management.

Adoption metrics after the December general availability, including the number of active teams and frequency of AI‑generated PR reviews.

Quality and error rates of AI‑generated code, especially any incidents that require extensive human rework or introduce bugs.

User feedback on the meeting‑agent beta, focusing on accuracy, privacy concerns, and impact on meeting efficiency.

Changes in Hops’ hiring practices and internal role definitions, which could indicate a broader industry move toward AI‑agent management.

Competitive responses from other collaboration platforms (e.g., Slack, Microsoft Teams) that may introduce similar AI‑agent features.

Related guides & quizzes

Found this useful?