Notes & AI

Capture engagement context and let AI surface insights across your organization.

Notes are the connective tissue of Servantium. They capture the raw context from every client interaction and feed the AI systems that power similar engagement matching, quote generation, and the institutional memory engine.

Creating notes

Notes are attached to engagements. Each note captures a single interaction or observation.

Open an engagement and navigate to the Notes tab.
Click Create to add a new note.
Enter the note text in the Note field. This is a free-form text area with up to 20 lines of visible space.
Optionally use the speech-to-text button (microphone icon) to dictate your note. Servantium uses the device’s speech recognition to transcribe your voice in real time. Click the microphone to start, speak, and click again to stop. The transcribed text is appended to any existing content in the note field.
The date is set automatically to the current time.
Click Create to save the note.

Tip

Capture notes as soon as possible after a client conversation. The details you record — technologies mentioned, budget signals, timeline constraints, decision-maker preferences — become the fuel for Servantium’s AI features.

Managing notes

The Notes tab displays all notes for an engagement as a chronological list. Each note card shows:

A preview of the note text (truncated to 20 characters for the display name)
The note date
Pin status

Pinning notes

Important notes can be pinned to keep them visible at the top of the list. Pinned notes surface key context without scrolling through the full history.

Speech-to-text transcription

Servantium includes built-in speech-to-text for hands-free note capture. This is especially useful for:

Recording observations during or immediately after a call
Capturing context while on the go (mobile app)
Quickly getting rough notes down that you can refine later

The speech recognition:

Initializes automatically when the note form loads
Appends transcribed text to existing content (does not overwrite)
Updates in real time as you speak
Works on any device with microphone access (requires browser/OS permission)

How notes power AI

Notes are far more than a record-keeping tool. They are the primary input for Servantium’s intelligence layer.

Note embeddings

When notes are saved, a backend trigger aggregates all notes for the engagement and generates a vector embedding — a mathematical representation of the engagement’s context. This embedding is stored on the engagement document and used for similarity search.

The process:

All notes for the engagement are collected and concatenated
The combined text is sent to Google’s Gemini Embedding model
The resulting vector is stored as notesEmbedding on the engagement
Firestore Vector Search indexes the embedding for fast similarity queries

AI context extraction

When an engagement has both notes and a template assigned, you can click the AI button (sparkle icon) on the Info tab to have AI read your notes and automatically fill in the template fields.

The AI:

Reads the engagement template’s field definitions (text fields, chip selectors, sliders, etc.)
Reads all notes for the engagement
Uses Gemini to extract relevant information from the notes and map it to template fields
Returns structured JSON that populates the engagement’s custom data fields

This turns unstructured conversation notes into structured, queryable engagement data.

Caution

AI context extraction is a suggestion, not a commitment. Always review the auto-filled values. The AI uses your notes as its source, so the quality of extraction depends on the detail in your notes.

Feeding similar engagement matching

The note embeddings enable the Similars tab on engagements. When you view an engagement’s Similars tab, Servantium calls a Cloud Function that performs a vector similarity search across all engagement embeddings in your organization. The most semantically similar engagements are returned, regardless of naming conventions or keyword overlap.

This means an engagement about a “Salesforce CRM rollout” can match against a past “Customer Relationship Platform implementation” because the underlying context is similar — even if the exact words differ.

Feeding AI quote generation

When you trigger AI quote generation, the system reads the engagement’s notes alongside data from similar engagements to build an accurate, context-aware quote. Notes provide the “what does the client actually need” context that makes AI-generated quotes meaningful.

The institutional memory engine

Notes are the foundation of Servantium’s institutional memory. The cycle works like this:

Capture — Team members record notes during and after client interactions
Embed — Backend triggers generate vector embeddings from the notes
Match — New engagements find similar past work through vector search
Generate — AI uses notes + similar engagements + the Service Catalog to build quotes
Learn — Completed engagements feed back into the system, making future matches and estimates more accurate

Each completed engagement makes the next estimate better. Notes are the input that starts this cycle.