Opslane

What do they actually do

Opslane currently ships an open‑source, local desktop app that lets developers run multiple Claude Code sessions in isolated Docker containers against a local Git repository. You describe a task, watch live diffs and logs as the AI edits files inside the container, and only sync changes back to your working copy when you approve them (site, GitHub README).

The app is local‑first and emphasizes explicit human control: multi‑session management, per‑session isolation, live diff previews, optional two‑way sync between the container and your repo, and session archiving (site, GitHub).

Separately, the team is exploring a SaaS “MCP” production ops control plane with connectors (e.g., Stripe, Supabase, Vercel) that shows previews, requires human approvals for writes, and keeps audit logs, and their YC listing markets “Deploy to Kubernetes in minutes.” Those are prospective directions; the shipped product today is the local desktop AI coding tool (opslane.dev, YC listing).

Who are their target customer(s)

• Individual developers who want to try AI edits without risking their working copy: They need a way to run experiments in isolation, see a clear diff of changes, and only apply edits to their repo when they’re ready.
• Small SaaS founders or ops people who need controlled production changes: They want read‑only access by default, explicit human approvals for any write, and audit logs for accountability when connecting to systems like Stripe, Supabase, or Vercel (opslane.dev).
• Engineers exploring multiple AI workflows or models at once: They need reproducible, isolated sessions with logs so experiments don’t interfere with each other or their local repo.
• DevOps/platform engineers responsible for deployments: They want faster previews, safer rollouts, and predictable rollbacks for Kubernetes deployments—the pain points referenced in Opslane’s YC positioning (YC listing).
• On‑call engineers who saw earlier Opslane alert‑classification tooling: They want less noisy alerts and faster triage to focus on real incidents (DEV post).

How would they acquire their first 10, 50, and 100 customers

• First 10: Personally onboard GitHub stargazers/contributors, YC contacts, and developer friends via 30–60 minute sessions; capture feedback to remove install friction and provide a one‑line installer and demo repo.
• First 50: Run live demos and post step‑by‑step tutorials in developer communities (GitHub Discussions, Dev.to, relevant Slack/Discord/subreddits) and invite Supabase/Stripe/Vercel founders to a closed beta; ask each new user for a referral and a short case note.
• First 100: Publish 1–2 short case studies and co‑market with priority connectors; add in‑app onboarding/templates, automate outreach to YC startups and developer newsletters, and run small targeted campaigns to the best‑converting communities.

What is the rough total addressable market

Top-down context:

Top‑down: the current desktop product sits in the AI developer tools market (~USD 5–6B, 2023–2024). The production ops control‑plane vision touches AIOps and DevOps markets (low‑ to mid‑double‑digit billions), while the Kubernetes delivery niche is ~USD 2B+ and larger within broader cloud‑native spend (AI code tools, AIOps, DevOps, Kubernetes, cloud‑native).

Bottom-up calculation:

Bottom‑up (desktop): ~19.6M professional developers in 2024, with ~62% using AI tools ≈ ~12M receptive users; if 5% pay $10–20/month, that’s ~$72–144M/year (JetBrains dev population, Stack Overflow 2024). Bottom‑up (ops control plane): assuming a reachable 100k SaaS teams paying $3k–15k/year implies ~$300M–$1.5B/year.

Assumptions:

• 5% paid adoption among ~12M AI‑tool users; price point $10–20 per developer per month for a local‑first desktop tool.
• Reachable segment of ~100k SaaS teams using systems like Stripe/Supabase/Vercel for the control plane; ARPA $3k–15k/year.
• Enterprise buyers for ops/Kubernetes products require security, approvals, and audit features, which support higher ARPA but longer sales cycles.

Who are some of their notable competitors

• GitHub Codespaces + Copilot: Cloud development workspaces plus AI coding assistance. Offers disposable, repo‑tied environments and editor‑integrated AI suggestions—overlapping with isolated AI iteration, but hosted and editor‑centric.
• Gitpod: Ephemeral, reproducible dev environments from your repository. Competes on isolation and session reproducibility, but via hosted workspaces and CI‑style automation rather than a local desktop orchestrating containerized AI sessions.
• Vercel (preview deployments & rollbacks): Fast preview builds and easy rollbacks for frontend apps. Overlaps with the “deploy previews/safe rollouts” angle, but focuses on frontend hosting rather than local AI edits or an ops control plane.
• Argo CD / Argo Rollouts: Open‑source Kubernetes delivery and controlled rollouts. Addresses safe deployments and rollbacks for K8s teams; infrastructure‑level CD tools rather than a human‑approved control plane or local AI editing surface.
• Teleport: Secure access and audit logging for servers, clusters, and databases. Aligns with least‑privilege and audit goals of the ops control plane, but focuses on access/authN/authZ vs. connector‑based, human‑approved writes across SaaS APIs.