AI for EdTech

Docently · EdTech AI module

A support agent for your learners and admins, trained on your help center.

Docently answers 'how do I' questions from learners, instructors and admins using your help center and docs, inside your product or in Slack, and hands off to your support team with the context attached when it can't.

A reference build, not software you license. It ships under your product's name.

admin.northstarcampus.com / help

Northstar Workforce Academy · Docently

Demo
  • How do I assign the safety course to everyone in the new Field Service department?

Step 1 of 5

Illustrative demo on a fictional institution. Not client work.

The problem

Every term start brings the same tickets: how to reset a password, where the gradebook export lives, why a course isn't showing. Support teams answer them by hand, and generic bots guess when the docs don't cover it.

The worked example

Northstar Workforce Academy (fictional). An employer admin asks how to bulk-assign a course to a new department.

Retrieval over your knowledge base with a fallback when nothing relevant is found, an embeddable widget and Slack integration, and routing to your existing help desk.

AI inside the workflow

Docently inside the product, not a chatbot beside it.

Switch between the product as it ships today and the same screen with the module working inside it. Ask a question, then approve what it prepares.

Northstar Learn · fictional product

Northstar Learn today: support answers the same how-to tickets by hand every term start.

No assistant

The product today. Every answer is reached by filtering, reading and clicking.

admin.northstarcampus.com / support / tickets

Northstar LearnDashboardCoursesLearnersSupportReportsAdmin
SupportOpen ticketsHelp centerEscalationsMacros

Open tickets

4 of 4 records

Bulk actions
Search
RefQuestionStatus
TKT-5501Bulk-assign a course to a departmentOpen
TKT-5503Reset a learner's passwordOpen
TKT-5506Export gradebook to CSVOpen
TKT-5509Different due dates for contractorsOpen

Search the help center, copy the steps, paste the answer, one ticket at a time.

The assistant runs Docently, branded as Northstar Help, the way it would ship under your product's name.

Illustrative demo on a fictional institution. Not client work.

What it does

What Docently does inside your product.

Best fit: LMS support, student services, K-12.

Trained on your docs

Ingests help-center articles, FAQs and release notes, and stays current as they change.

Embeddable

A widget in your product, a Slack app for internal teams, or both.

Knows its limits

Says when the docs don't cover a question instead of guessing.

Clean handoff

Opens a ticket with the conversation and the user's context already attached.

Where it comes from: Built on the support-agent pattern we've shipped twice.

See a support agent we shipped

How it works

One AI feature, working inside your product, in two weeks. $10,000, fixed.

In two weeks: Docently answering from your real help center, tested on a set of past tickets, with the share it could answer and the cost per conversation.

  1. 01

    EdTech Product Review

    30 minutes, free

    We look at your product with you and pick the module or capability that protects your next release, what it would cost to serve, and what your customers' security review will ask.

  2. 02

    The Sprint

    $10,000, two weeks, fixed

    One AI feature working inside your product: running in your staging environment on your catalog, documents or synthetic data, behind a feature flag, delivered as pull requests. You also get quality and safety evals, a cost estimate on two models, a data-flow sheet and a fixed-price quote for production.

  3. 03

    Targeted Unlock, then Capability Waves

    Quoted at the end of the Sprint

    The first production capability, delivered by pull request into your repo with no standing production access. Then the next module, on the foundation the first one built.

Built for education data

Controls your customers' security reviews ask about.

Each control is a design fact about how the module is built and how we work, not a compliance claim.

  • Runs on your catalog, documents or data through the APIs your product already exposes. No rebuild.
  • Sprint work uses synthetic or de-identified data and ships as pull requests to your repo. No production student data.
  • Every answer that matters shows its source; material actions go to a person for approval.
  • Built around FERPA, COPPA, accessibility and institutional requirements.
  • Built to ship under your product's name. Model and hosting are your choice.

We build around your FERPA, COPPA, accessibility, security and institutional requirements. Compliance depends on the complete product, contracts, policies and operating environment; Datics does not present software engineering alone as legal certification.

How we handle student data

Work this builds on

Projects we have delivered that use the same engine or the same engineering pattern.

AI Knowledge and Support Agent

Cloud Storage Software · Support Agent / Knowledge Retrieval

An AI agent that answers from knowledge spread across cloud storage, inside Slack and Teams

AI that ingests and indexes documents and pages across several cloud-storage providers and answers in Slack, checking approved answers first and falling back cleanly when it has nothing relevant.

Support agentRetrievalSlack
Read case study

Cited Knowledge Workspace

Knowledge & Document Software · Knowledge Retrieval / Cited Answers

Answers grounded in an organization's own documents, with every source cited

A self-hosted workspace that extracts documents, answers questions with citations, respects who may see what, and helps people write.

Cited answersRetrievalSelf-hosted
Read case study

Document Intake Assistant

Financial Advisory · Document Intelligence / Intake Automation

Documents read into the record and multi-step forms pre-filled, with access kept separate

An assistant that reads the decks, notes and documents fund managers submit, pre-fills multi-step onboarding and due-diligence forms, scores them and keeps each user type to its own data.

Document AIIntake automationAccess control
Read case study

See Docently on your own product.

Bring one product, workflow or AI initiative. In 30 minutes we pick the module that fits first, what it would cost to serve and what your customers' security review will ask.

Calls run in US Central time. No generic AI presentation: we discuss your product, users and production constraints.