# Vinaka.AI

> The investigation platform for B2B support teams. Investigate incidents collaboratively, preserve every decision, and let AI recommend the next best action using your historical resolutions. One platform for all investigations — from first alert to final resolution — on a Resolution Graph, not keyword search over your wiki.

## What Vinaka does

Four workflows in one platform:

- **Investigate.** AI reconstructs every investigation using historical knowledge — findings, similar incidents, and a recommended approach before the queue opens.
- **Collaborate.** Bring support and engineering into one shared platform. Support, Engineering, and AI work in one owner-attributed session.
- **Track.** Nothing gets lost during handoffs or escalations. Every action, escalation, status change, and decision is preserved on the ticket.
- **Reply.** Generate contextual customer updates backed by evidence. Draft, review, and send cited customer replies from Vinaka today.

Vinaka can investigate incoming tickets proactively: analysis, sources, and recommendations land in the thread before your engineer opens the queue. Available in Phase 2 early access.

## Who it's for

For B2B support teams, L1 and L2 through TSE. Not a chatbot. Not an AI SRE.

Software companies where tickets require deep product and environment knowledge and tribal knowledge is expensive to lose.

Built by engineers from Cohesity, the team support used to escalate to.

## Status

The product is built and running on real data. We're selecting our first design partners now.

## Integrations

Plugs into your stack. Replies go out from here.

- **Jira** — ticket history, comments, and engineering context for matching and escalation (read-only OAuth)
- **Confluence** — runbooks, procedures, and architecture docs surfaced in investigation (read-only OAuth)
- **Slack** — resolution patterns and tribal knowledge from the threads where fixes actually happen (read-only OAuth)
- **Salesforce** — account context, entitlements, and customer environment data (read-only OAuth)
- **Customer replies** send from Vinaka today
- **Source write-back** to Jira/Salesforce: opt-in, on the roadmap
- **Planned connectors:** Zendesk, ServiceNow, PagerDuty

Data: read-only OAuth by default, scoped per integration. You choose what gets indexed. Permissions follow your existing access controls. We never train on your data. Vinaka suggests; your engineer decides. No autonomous actions on your systems.

## Product primitives

- **Resolution Graph**: structured investigation memory from every closed ticket — evidence, environment, decisions, and outcomes.
- **Investigation state**: the tuple of what's been tried, ruled out, and found for a given ticket. Future tickets are matched against investigation state rather than keywords.
- **Calibrated confidence**: a score attached to each recommendation, backed by past-resolution reuse data.
- **Proactive mode**: recommendations prepared before the engineer opens the ticket (Phase 2).

## Canonical pages

- [Home](https://vinaka.ai/): product overview, platform pillars, Resolution Graph.
- [Platform](https://vinaka.ai/#platform): Investigate, Collaborate, Track, Reply.
- [Resolution Graph](https://vinaka.ai/#demo): how closed tickets compound into investigation memory.
- [FAQ](https://vinaka.ai/#faq): integrations, data handling, proactive mode, onboarding.
- [Integrations](https://vinaka.ai/#integrations): connectors and data access model.
- [LLM summary](https://vinaka.ai/llms.txt): this file.

## Not Vinaka

- Not a generic GPT wrapper or flat wiki RAG.
- Not a ticket deflection / chat-first consumer support tool.
- Not an AI SRE for production incidents.

## Contact

- Early access: https://vinaka.ai/#cta
- Email: hello@vinaka.ai