Contact

00 Independent technology initiative

Engineering intelligence
for the real world

Building intelligent products today, with a path toward advanced automation, autonomous systems and robotics.

Products
TaskMatch · Agli Tareekh
Discipline
Applied AI systems
Direction
Autonomy & robotics
Kinematic solver · 3-link planar Tracking
AXIS θ J1 — J2 — J3 — WORKSPACE ENVELOPE
SolverCCD · 10 IT Error StateNOMINAL

01 Products

Developing software
Android first
Multilingual by design

Two products, one idea: remove the friction between people, information and action.

Both products start from the same observation — most software fails people at the moment they are least able to fight it: an unfamiliar language, an unclear agreement, a date they cannot afford to miss.

Product 01 · Android · in development

TaskMatch

Voice-first task marketplace

A multilingual marketplace connecting people who need work done with people who can do it. Designed around spoken interaction, eleven languages, and users who should never have to read a form to be understood.

  • Design ruleVoice-first, never voice-only — every step is designed to have a touch fallback
  • LanguagesEleven planned: English, Arabic, Urdu, Hindi, Bengali, Malayalam, Tamil, Telugu, Nepali, Filipino, Pashto
  • Core ideaAn agreement both sides can see and understand before work begins
S1

Speak

The task is described out loud, in the speaker's own language.

S2

Understand

Language, intent and scope are resolved into structured detail.

S3

Match

Nearby capability is paired with the task and its constraints.

S4

Agree

Price, scope and arrival are confirmed by both sides and recorded.

S5

Complete

Changes stay visible, and the finished record belongs to both parties.

Product 02 · Android · in development

Agli Tareekh اگلی تاریخ

Legal companion for court practice

A hearing diary for litigants and lawyers. Leave the courtroom, record the next date quickly, and keep the case's current position in one place instead of scattered across notebooks and phone reminders.

  • Built forDistrict and high court practice, starting in Pakistan
  • Offline firstBundled court lists, local records, password-encrypted backup and restore
  • Two profilesPersonal case tracking, and a lawyer diary for a caseload
Hearing railILLUSTRATIVE
14 MAR · HEARD ADJOURNED — 1 02 APR · HEARD ADJOURNED — 2 21 APR · NEXT DATE ATTENDANCE REQUIRED STAGE · EVIDENCE NOT SCHEDULED
Current positionSAMPLE DATA
Next date21 April · 09:30
StageArguments — civil suit
Action neededFile reply, bring witness list
Documents2 of 3 prepared
Client informed Yes — 14 March
ReminderEvening before, in Urdu

02 Core technology

Shared foundations
across every
AlimAI product

The parts we build once and reuse everywhere.

01

Multilingual interaction

Speech and text handled in the language a person actually thinks in, including scripts and reading directions that most interfaces treat as an afterthought.

02

AI-assisted workflows

Models sit behind the interface, not in front of it. They clarify, structure and draft — the person decides, and the decision is what gets recorded.

03

Contextual continuity

State that survives between sessions: what was agreed, what changed, and what is still open — kept per task and per case, never pooled across them.

04

Intelligent automation

Routine steps — reminders, follow-ups, status changes — run on their own, with a clear boundary around what a system is allowed to do unattended.

03 Research

Exploratory work
Independent
Ongoing

Products raise questions that products cannot answer.

AlimAI keeps a small research line because the hard part of intelligent systems is not generating text — it is holding a correct, shared understanding of a situation over time. That question sits underneath every product here, and it becomes harder, not easier, once systems begin to act in the physical world.

R-01

Intelligent state modelling

Representing what a system currently believes about a task, and what would have to change for that belief to be wrong.

R-02

Contextual continuity

How understanding is preserved, degraded or lost across sessions, devices and long gaps in interaction.

R-03

Human–AI interaction

Where a system should ask instead of assume, and how it shows its reasoning without burying the person in it.

R-04

Observer-specific representation

The same event described differently to different participants, without either description becoming untrue.

This work is exploratory. Research ideas inform how AlimAI products are designed, but they are not presented as validated technology, and nothing here is offered as a proven result.

04 Robotics & future systems

Software first. Then systems that sense, decide and move.

The same problem repeats at every scale: a system has to build an accurate picture of a situation, keep it current, and act on it without losing the thread. Solving that in an app is the groundwork for solving it in a machine.

AlimAI treats robotics as a destination reached in order — agents, then automation, then perception, then physical autonomy. Each stage has to earn the next.

Future direction — no robotics product released
ACTUATION CONTROL LOOP SPATIAL MODEL PERCEPTION FIG. A — CONCEPT STUDY NOT A PRODUCT SPECIFICATION
STAGE 01In development

AI agents

Assistants scoped to a single task or case, drafting and clarifying under human approval.

STAGE 02Planned

Intelligent automation

Routine sequences that run unattended, with explicit limits on what may change without review.

STAGE 03Exploratory

Sensing & perception

Turning sensor input into a spatial and situational model a system can reason over.

STAGE 04Long-term

Autonomous systems

Machines that plan and act in the physical world, and stay accountable for what they did.

05 About

Khyber Pakhtunkhwa, Pakistan
Independent
Engineering-led

An independent engineering initiative with a long horizon.

AlimAI is an independent technology venture developing its own products and conducting independent research.

The near-term work is software: multilingual interfaces, AI-assisted workflows, and systems that keep a reliable record of what was decided. The longer-term intent is to move that competence toward machines that sense and act — deliberately, and only as each stage is actually earned.

Everything on this site describes what exists today or what is being explored. Where something is a direction rather than a product, it says so.

Alim ul Haq Khan Founder · AlimAI

06 Contact

Partnerships
Collaboration
Strategic discussion

Start a conversation.

AlimAI is open to work that makes the products stronger or moves the longer-term direction forward. Direct enquiries are read and answered personally.

  • ATechnology partnerships and integration
  • BProduct collaboration and pilot deployments
  • CResearch collaboration
  • DStrategic and investment discussions

Direct contact

Message on WhatsApp for a fast reply, or email for anything with documents attached.