• As we may think.
  • As we may learn.
  • As we may grow.
  • As we may flow.
  • As we may create.
AS WE MAY

AI systems that remember a company’s context, connect its knowledge and turn it into the next step.

Our solutions
1945 → nowScroll ↓

AI systems with a memory for context

Human intelligence, extended.

We gather a company’s context from meetings, chats and files, connect it to projects and turn it into a decision you can open and act on.

Our solutionsBook a demo

Approach

AI should not start from a blank window.

Most systems today start over every time. We build products that keep a company’s context and use it to help take the next step.

  1. 01

    Remember

    Keeps the record

    What a person knows, what they did, where they went wrong and what they agreed on.

  2. 02

    Understand

    Sees the links

    How it all connects to the company’s projects, roles and goals.

  3. 03

    Act

    Moves it forward

    Which next step is needed and why this one — with a link to the source.

An idea, continued

1945

As We May Think

Vannevar Bush’s essay. He described the Memex — a machine that keeps everything a person has read and decided, and links it with associative trails. Technology lagged behind the idea by eighty years.

now

Now we may.

Not because models appeared. Because it finally became possible to hold a company’s context longer than a single conversation.

1945196519852005now

Memory became context.

Context became intelligence.

Intelligence is becoming action.

Solutions

Four endings to one sentence. The verb comes first, the name second — so it reads as one system, not a catalogue of unrelated services.

As we maythink

AI that pays for itself on a real task.

AI adoption for organizations: we find the processes where AI delivers measurable results, build tools and multi-agent systems for them inside your perimeter and prove the impact in numbers.

TaskBaselinePilotImpact
Explore Think
Document preparation from a templateSample · illustrative data · Prototype · week 3

Baseline before the start

4 h per task30 tasks a monthThreshold −40%

Pilot progress

  • 01A multi-agent system built on the company’s data.Week 1
  • 02Pilot launched with a team of 8.Week 2
  • 031 h per task instead of 4: threshold passed.Week 3 measurement

Next step

If the measurement clears the threshold, agree on scaling to a second department; estimated impact: 90 h a month.

As we maygrow

A company knows its people’s job titles. But does it know what they can do?

Skill intelligence for organizations: roles, skills, verified levels, gaps — and how ready the team is for the work still ahead.

RolesSkillsVerificationGaps
Open grow.aswemay.digital
Team — Product developmentSample · illustrative data · 14 people

Role, skill, verified level

  • Requirements managementAnalystready
  • Working with dataAnalystgap 2
  • Release planningTeam leadgap 2

Next step

Two people close the data gap with one course — faster than hiring.

As we maylearn

A company’s knowledge should not stay locked in documents.

Documents, expertise and goals become a structured course with practice and checks — learning people can apply next week.

MaterialsStructurePracticeChecks
Open learn.aswemay.digital
Course — Working with requirementsSample · illustrative data · Draft assembled

Source materials

  • Requirements_approval_policy.pdf
  • Rollout review at a logistics company
  • Expert interview, 40 min

Course outline

  1. Module 1 — where the process breaks most often
  2. Module 2 — how to decide on a real case
  3. Practice — a situation from your own company
  4. Check — 8 questions on application

As we mayflow

Working memory that never loses context.

Meetings, voice notes, tasks and agreements gather into one memory and come back when the context is actually needed. Short meetings are processed on your computer; long ones, on a processing server inside your perimeter.

HearsUnderstandsConnectsReturns
Open flow.aswemay.digital
Meeting 14:03 — architectureSample · illustrative data · 4 items to review

What was understood

  • decisionWe keep the second architecture option.
  • taskMisha checks the API by Friday.
  • questionWhich tools go into the first scope?

Linked

Project · flowMisha · APIMeeting 24.08

Next step

Before Thursday: follow up with Misha on the API deadline — promised by Friday.

Different products. One idea.

Different tasks. One principle.

As we may

think.

AI adoption

For companies

passes to grow.which skills the new processes need

grow.

Organization

For the organization

passes to learn.what to teach, and to whom

learn.

Knowledge

For L&D

passes to grow.the verified level after the course

flow.

Working memory

For leaders

passes to think.where the team loses time and decisions

CONTEXTNEXT STEP

The systems get more useful as they understand the organization’s context better.

As we may become.

The sentence is left unfinished on purpose. It is completed not by the system but by a person — through what they do next.

AS WE MAY

Next step

What could AI change in your work?

Tell us about your task — we will show a fitting solution or discuss a pilot. We reply within a day.

Data processing consentPrivacy policy

Contact us
Format
Diagnostic: 2 weeks; prototype: by week 3; pilot: 6–10 weeks; then a scaling decision.
Deployment
Think, Grow and Flow run inside your company’s perimeter. Learn and Lab use an external LLM; their deployment terms are discussed separately.
Who to talk to
Dmitry Krechman, business development director — dimak@aswemay.ru