Company brain · live installation

Meet Novi. The AI teammate with your whole company in her head.

Novi is a deployed, working business agent. She reads where your team already works — Google Chat, Meet, Loom, Drive, Telegram — remembers with permission awareness, answers with citations, and acts: tasks, projects, durable memory, live operational data, images, web research, and delegation to engineering agents. All through one governed brain.

30+guarded tools
6ingestion sources
100%tool calls audited
0uncited answers
Business memory layer Memory + tools + operational data + permissions Chat + Telegram DMs, channels, voice Meet + Loom transcripts, actions Drive + SOPs docs, reviewed notes Ops + web live read-only tools Brain API auth, ingest, retrieval, tools, audit, promotion Novi Chat + Telegram MCP Agents Vault Humans Core rule Filter by requester and audience before the LLM sees context.

The model is not the product. The governed operating layer around it is.

The actual problem

Most business AI feels shallow because company context is fragmented.

Chat threads, meetings, SOPs, dashboards, vendor notes, Drive docs, and operational databases all hold pieces of the truth. The company brain gives Novi and connected agents a governed way to remember, cite, and act on that context.

Operational memory

What happened

Raw messages, meeting transcripts, Loom and Telegram voice transcripts, source links, action items, image metadata, reminders, turn logs, and audit trails live in Supabase.

Durable knowledge

What the company decided

Reviewed SOPs, decisions, project summaries, recurring workflows, and lessons become indexed notes — with governance tiers up to company rules and laws — then get promoted into a human-readable GitHub/Obsidian vault.

Live tools

What is true right now

Novi can use guarded tools for tasks and projects, Drive search, thread summaries, live operational systems, public web search, image flows, her own codebase, and delegation to engineering agents.

What Novi does all day

A general business assistant, not a search box.

Ask in plain language, in the tools your team already uses. Novi figures out which of her capabilities the request needs, checks permissions, and does the work.

Answer

"What did we decide about the Q3 supplier?"

Permission-filtered hybrid search over everything ingested — chat, meetings, Looms, docs, notes — with citations on every answer and four retrieval depths from cheap to deep.

Act

"Add a task, mark it done, start a project."

Action items are created and updated instantly and audited. Projects and durable memories are drafted first, then created only after you confirm the exact record.

Check live systems

"Are any fulfillment orders blocked?"

Read-only connectors into operational Baserow and MySQL answer with current state — sync health, app errors, order status — always stamped "checked live at" and never passed off mock data as real.

Remember

"Novi, remember this."

Novi drafts the exact durable note — title, body, type, tags, visibility — and saves only after you confirm. Notes are searchable immediately and flow through review governance to the curated vault.

See & create

"What's in this screenshot?" "Make me a diagram."

Images posted in Chat or Telegram are stored privately and used as vision context. Novi can repost any stored image you're allowed to see, or generate new images and post them back to the conversation.

Delegate

"Ask the engineering agent to build it."

Novi hands scoped software tasks to registered engineering agents (e.g. Hermes), which work autonomously and open a PR for human review. High-risk work is held for admin approval before it starts.

Watch & follow up

Nobody falls through the cracks.

If someone is @-tagged in a channel and hasn't replied in 24 hours, Novi posts one polite, batched nudge. An emoji reaction counts as a reply; "cc/fyi" tags never nag; history imports never trigger pings.

Detect work passively

Assignments get noticed, never auto-filed.

Novi can scan normal channel conversation for assignments and resolutions, then ask in-thread: "Track this as an action item?" Nothing is created until a person clicks confirm — and the click is re-authorized server-side.

Explain herself

"Novi, how do you handle permissions?"

Novi can search and read her own source code and docs, answer with commit-pinned permalinks, and even file a scoped change request for her own behavior — after you confirm the spec.

Architecture

The Brain API is the gatekeeper.

Google Chat and Telegram are the user surfaces, and the same Brain API and MCP connector expose controlled access to external agents. No side agent gets database credentials or bypasses permission checks.

SYSTEM SHAPE Sources flow into one permissioned brain. Ingestion sources Google Chat + images Google Meet transcripts Google Drive and SOPs Loom transcripts Telegram text + voice Manual compliant imports Live ops connectors raw first, LLM later Cloud Run Brain API auth, ingest, answer, tools, audit RequestContext guard Permission-filtered retrieval Novi tool registry (30+) Promotion and governance Supabase + pgvector messages, chunks, notes, actions, images, audit Novi in Chat + Telegram DMs, channels, threads, voice, tools MCP and agents Claude, Codex, Hermes, OpenClaw GitHub/Obsidian vault reviewed durable mirror for humans
Novi

The chat agent is a planning tool loop, not a single-shot chatbot.

Every DM or mention runs the full agentic loop: plan, pick a tool, pass its permission guard, execute, observe, repeat as needed, then synthesize a cited answer. Novi posts a 🤖 working indicator instantly and edits it in place when the answer is ready, so slow multi-tool turns never time out.

NOVI LOOP Plan, guard, execute, cite. User asks DM or channel Plan choose tools Tool call requested by LLM Permission guard guarded per tool Execute safely search, act, Drive, ops, web Observe result safe summaries and citations Answer grounded, cited, audited
Built like a teammate

Real conversation memory: Novi stores both your turns and her own replies per conversation, so "what did you just say?" and "what about that other customer?" work like talking to a person.

A personality: a warm, sharp coworker with opinions — leads with the answer, brief by default, dry humor when it fits, flags risks early, and says "I can't see that" plainly. Tone tightens automatically in shared channels.

Quality-gated: offline agent evals and a quality gate run before any prompt, tool, or retrieval change ships; every turn writes a sanitized, gradeable trace for diagnostics.

Permission model

DM answers use requester access. Channel answers use requester access intersected with audience access. Private DMs, sensitive channels, protected images, and restricted Drive files never leak into broader answers — and tool selection by the AI is never authorization: every tool runs its own server-side guard before executing.

The full tool catalog

Every tool Novi can pick up — all guarded, all audited.

This is the complete shared registry behind Novi in chat and the ask_novi MCP surface. Read tools are always available; write and action tools are offered only to callers allowed to write, and several require an explicit draft-then-confirm step or admin authority.

Knowledge & recall

Memory, threads, and the public web

search_brain
read

Permission-filtered hybrid search over chat, meetings, Looms, docs, and notes — with conservative, balanced, deep, and legacy semantic modes and citations on every result.

summarize_thread
read

Key points, decisions, and open questions from the current Chat thread, Telegram DM, group, or forum topic — only messages the audience may see.

search_web
read

Grounded public web search for current or external facts the brain can't answer — labeled, cited, freshness-stamped, and treated as untrusted data that never overrides company memory.

Tasks & projects

Getting work tracked and done

list_action_items
read

Your open tasks, filterable by status, with private items withheld from channel answers (and a nudge to DM for them).

create_action_item
write

Adds a task instantly — title, details, due date, priority — inheriting the visibility of where it was created.

update_action_item
write

Mark done, reopen, reschedule, rename, reprioritize, or add resolution notes — by id or just a distinctive part of the title.

create_project
writedraft → confirm

Drafts a first-class project record (including from a pasted pod-profile) and creates it only after you approve the draft.

Durable memory

Capture, governance, and supersession

remember_memory
writedraft → confirm

The "@novi remember this" flow: drafts the exact note — title, body, type, tags, confidence, visibility — and saves only after explicit confirmation.

govern_brain_memory
write

Moves a note through review: request review, review, accept, reject, or deprecate. Admin-only transitions are enforced server-side.

supersede_brain_memory
writeadmin

Replaces an accepted fact, rule, SOP summary, or decision with a new note and marks the old one superseded — truth changes by replacement, never silent edits.

detect_memory_conflicts
write

Finds documents, messages, or accepted notes the candidate may duplicate or contradict, and files reviewable conflict rows before anything is accepted.

resolve_memory_conflict
write

Marks a conflict resolved or dismissed after human review — required before accepting a candidate with open conflicts.

propose_memory_change_request
write

Lets any team member propose a change to accepted knowledge instead of editing it directly.

resolve_memory_change_request
writeadmin

Accept, reject, or cancel an open change request.

apply_memory_change_request
writeadmin

Applies an accepted supersession request through the same governed replacement path, preserving who approved it.

promote_to_rule
writeadmin / lead

Elevates reviewed knowledge to rule tier — higher-authority memory that ranks above ordinary notes.

set_law
writeadmin

The highest tier: company law. Feature-flagged, admin-only, fully audited.

Documents & Drive

The files your team actually works in

search_drive
read

In a DM, searches Drive as you — results are exactly the files you can open. In a channel, searches only admin-mapped folder roots so a requester's extra access can't leak files to the room.

draft_channel_drive_mapping
writeadmindraft → confirm

Drafts and, after explicit approval, activates the Drive folder roots a channel may search.

list_channel_drive_mappings
readadmin

Lists active and disabled Drive roots mapped to each channel.

set_channel_drive_mapping_state
writeadmin

Disables or re-enables a mapping without deleting its audit history.

Live operations

What is true in the business right now

query_baserow
read

Manufacturing and fulfillment orders straight from operational Baserow — current production and fulfillment state, stamped "checked live at".

query_mysql
read

Application state from MySQL: data-sync health, recent app errors, and customer or vendor status — read-only, never stale chat summaries.

Images & vision

Seeing and creating, inside the permission boundary

send_brain_image
action

Pulls up a previously stored image the requester and audience may access and posts it back into the current Chat or Telegram conversation.

generate_image
action

Generates a new image from your prompt, stores it under the conversation's visibility, and posts it back — no public URLs, ever.

Delegation & self-knowledge

Novi as a working colleague — and her own expert

delegate_task
action

Hands a software task to a registered engineering agent that works autonomously and opens a PR for human review. High-risk work (deploys, migrations, credentials) is held at an approval gate.

approve_delegation
actionadmin

Releases a held high-risk delegated task — admin-only, refused for anyone else.

search_codebase
read

Searches Novi's own source repository to answer "how do you work?" questions with commit-pinned links.

read_codebase_file
read

Reads exact files and line ranges from her own code and docs, citing immutable line-anchored permalinks.

request_codebase_change
actiondraft → confirm

Scopes a change to Novi's own behavior into a structured spec, shows you the draft, and files it to the engineering agent as a GitHub issue only after you agree.

Also on the MCP surface

Connected agents and IDEs get additional direct-contract tools through the per-user connector: ask_novi (the full loop), answer_from_brain, write_brain_memory (immediate indexed write for trusted callers), promote_to_curated_note, and create_reminder — all scoped to the calling user's token.

How to read the badges

read retrieval only   write creates or changes records, audited   action outward-facing effect   admin server-enforced authority   draft → confirm a human reviews the exact artifact before it exists.

Proactive Novi

She doesn't just answer. She notices.

The brain runs background workers that watch for dropped balls and turn raw activity into structured work — always with a human in the loop before anything binding happens.

Follow-up nudges

Unanswered tags get one smart bump

Tagged and silent for 24h? Novi posts one batched, @-pinging nudge per thread. Emoji reactions count as replies, "cc/fyi" runs are exempt, and history imports never trigger pings.

Passive action detection

Assignments become proposals

Novi can scan normal conversation for assignments and resolutions and ask in-thread before tracking anything. The confirm click is re-authorized server-side; nothing is auto-filed.

Meeting followups

Notes land where the team works

Company-visible meetings can post their summary and action items to a named Chat channel automatically — deduped per meeting, never leaking invite-only content to broader rooms.

SOP resync

Docs stay current on a schedule

Scheduled Drive SOP resync keeps ingested procedure docs aligned with the live files, so answers cite the version your team actually follows.

Durable memory

Raw history becomes durable knowledge only after review and promotion.

The vault is not the retrieval database. Supabase is the searchable source of truth; the vault is a human-readable GitHub/Obsidian mirror of accepted, promoted notes.

DURABLE MEMORY LIFECYCLE Ingest, index, review, promote, supersede. Source enters Chat, Meet, Loom, Drive, Telegram Store raw messages, docs, images, provenance Chunk + embed content_chunks visibility inherited Review note draft, proposed, reviewed, accepted Promote to vault GitHub-backed markdown mirror + rule and law tiers Change by supersession Accepted/promoted truth is not silently rewritten. Users create memory mid-conversation. "@novi remember this" drafts the exact note, waits for confirmation, then creates and indexes it under the current DM/channel visibility. Conflicts with existing knowledge are surfaced for review first.
Manual capture

During the conversation

Novi drafts title, body, memory type, tags, source summary, and inherited visibility. The user confirms before anything becomes durable. A bare /novi remember opens a form, same review flow.

Document ingest

From Drive and SOPs

Drive/SOP text is ingested into document rows and indexed chunks, with scheduled resync. The editable SOP stays in Drive; durable notes cite and summarize it.

Governed change

Conflicts, requests, tiers

New candidates are checked for conflicts against existing knowledge. Anyone can file a change request; admins accept and apply via supersession. Rules and laws sit above ordinary notes as higher-authority memory.

MCP and external agents

The Brain is not only for chat.

The same governed brain powers IDEs, coding agents, and dashboards. A personal token represents the user — it can only read or write what that user can.

External access

Per-user MCP tokens

Claude, Codex, Cursor-style tools, Hermes, OpenClaw, and dashboards connect through the Brain API or the published brain-connector MCP server. One npx command and two env vars and your agent has the company brain.

  • ask_novi — the full plan → tools → cited-answer loop
  • search_brain and answer_from_brain
  • write_brain_memory and promote_to_curated_note
  • list_action_items, create_action_item, update_action_item
  • create_reminder and summarize_thread
Delegation

Agents can work, but not free-run

Novi hands scoped work to registered agents such as Hermes. High-risk actions like production writes, merges, migrations, credential changes, or customer-visible changes require approval gates and audited callbacks.

Allowed pattern

Chat request -> Novi -> delegated task -> connector -> status callback -> Novi updates the thread with the PR link.

Not allowed

External agents do not connect directly to the database, scrape protected images, or bypass Brain API permissions.

Inputs and monitoring

The brain listens where the team already works.

Google Chat is the primary interface and Telegram is live alongside it — including voice. Meetings, Looms, Drive docs, and compliant manual imports round out the record.

Google Chat

Primary surface

DMs work instantly for org users — no admin setup. Channel mentions, passive ingestion, history backfill, images as vision context, thread history, and follow-up nudges.

Meet + Loom

Meetings and videos

Transcript ingestion with action-item extraction. Meetings default to invite-list visibility; a company-wide tag makes them company-visible, optionally posting summaries to a named channel.

Drive + SOPs

Working documents

Metadata-first discovery, explicit document ingest, scheduled SOP resync, and reviewed durable notes that point back to the source file.

Telegram

Groups, DMs, and voice

Approved groups and linked DMs are logged; photos flow through the same protected image pipeline; voice notes are transcribed and indexed as text. Mention @novi and she answers right there.

What makes it team-grade

A personal memory system asks, "Can my agent remember?" A company brain asks, "Can the system remember the right thing, show it to the right people, hide it from the wrong people, cite where it came from, and let multiple humans and agents use it without turning permissions into a mess?"

Runtime

Cloud Run API and workers, GitHub Actions continuous deploys gated on build + tests, Secret Manager-backed credentials.

Database

Supabase Postgres with pgvector and deny-by-default row-level security; content chunks, brain notes, actions, projects, audit logs, image metadata, and visibility policies.

LLM layer

Model-routed planning and synthesis, embeddings, versioned prompts, per-call token/cost logging, offline agent evals, and a quality gate before rollout.

User interface

Novi in Google Chat and Telegram: instant working indicator, async delivery, durable multi-turn conversation memory, and cited answers.

Human vault

GitHub/Obsidian mirror for reviewed durable notes, compiled SOP/wiki pages, source manifests, and a "request a change" path for readers.

Safety

Permission filtering before model context, per-tool guards, full audit records, draft-then-confirm gates for memory/projects/code changes, admin approval for high-risk delegation, and sanitized traces — no secrets, raw URLs, or private identifiers ever reach logs or answers.

← Back to the full portfolio