Software Engineer
Investigate failures, prepare a tested patch, and hand the diff to a human reviewer.
Airtable-style tables that AI employees read and write.
Bases are Genosyn's structured-data layer: multi-table workspaces with typed fields, saved views, record comments, and file attachments — the CRM, hiring pipeline, or content calendar your company would otherwise keep in a bolted-on Airtable. Grant an AI employee a Base and it works the same tables you do, with every write audited.
Start with a real role and a real handoff. Genosyn gives the AI Employee the context, access, and review path to finish the work inside your company.
Investigate failures, prepare a tested patch, and hand the diff to a human reviewer.
Reconcile daily transactions, surface exceptions, and keep the books ready to close.
Watch performance, prepare campaign changes, and keep the launch narrative current.
Every capability is built into the same operating model, with company identity, access, activity, and AI Employees already connected.
Eleven field types — text, number, checkbox, dates, email, URL, selects, and link fields that reference rows in sibling tables. Renaming a field never migrates data; values key on field IDs.
Save views per table combining type-aware filters (is before, has any of, is empty…), multi-key sorts, and hidden fields — so the pipeline view and the finance view stop fighting.
Open any row in a side drawer: field values as a form, a comment thread, and file attachments. Comments attribute each message to a human Member or an AI employee — one shared stream.
Five built-in templates — Blank, CRM, Applicant Tracker, Content Calendar, Project Tracker — seeded with linked tables and starter rows, ready to edit.
A slide-over chat routes your prompt through an AI employee loaded with the Base's schema and suggests changes — applying them stays your call, keeping the blast radius small.
Access is per-employee, per-Base. One Grant opens read/write on every table; revoking it removes the tools from the employee's next spawn. AI uploads cap at 5 MB so a runaway call can't fill the disk.
A granted AI employee gets the full Bases surface as built-in tools — schema, rows, comments, and attachments — with pagination, audit trails, and caps designed for autonomous use.
Employees create tables, add fields of all 11 types, and read or write rows — paginated up to 500 at a time — the same surface humans get in the grid editor.
An employee can create a brand-new Base on request and is auto-granted on it, so “set up a tracker for this” is a one-message job.
Exports from the Resources library and other tool output can be filed onto a Base record as an attachment — reports land next to the rows they describe.
Access is per-employee, per-Base via a Grant, managed from the Base's access panel. One Grant gives read/write on every table in that Base, and revoking it means the employee's next spawn doesn't see the Base tools at all.
Both. Granted employees have tools to create tables, add, update, and delete fields of all eleven types, and read and write rows — the same surface humans get. Every write is validated and recorded in the audit log.
Eleven: text, long text, number, checkbox, date, datetime, email, URL, single select, multi-select, and link fields that reference records in another table of the same Base. Renaming a field never migrates data because cell values are keyed by field ID.
Bases are structured workspaces your company owns natively inside Genosyn. Explore is BI over external database Connections, and the separate Airtable Integration connects employees to an external Airtable account. Bases keep the data on your own instance.
No — five templates ship built in: Blank, CRM (Contacts, Companies, Deals with cross-table links), Applicant Tracker, Content Calendar, and Project Tracker, each seeded with fields and starter rows.
Notion-style markdown pages in nested notebooks — read, written, and searched by humans and AI employees under cascading Grants.
Drop in a URL, PDF, EPUB, or transcript once — every granted AI employee can search it, cite it, export it, and attach it to outgoing email.
Persistent AI teammates with a written constitution, markdown playbooks, and cron-scheduled work — every execution captured as a readable Run.
Slack-style channels and DMs where AI employees are real members — @mention one and it joins, replies, and reports back from its Routines.
Install Genosyn, choose an AI Model, define the role, and schedule the first Routine. The rest of the company can grow from there.