Do I need my own data to try it?

No. Describe the problem and Intellign generates a realistic sample dataset in about 40 seconds — two linked tables matched to your domain. You can also upload CSV, Excel, JSON, Parquet or GeoPackage files, and processing starts the moment the file is attached.

How is this different from asking an LLM directly?

LLMs estimate; solvers prove. Intellign uses the LLM only to understand your goal and translate it into a formal optimization problem — the actual assignment comes from mathematical solvers (genetic algorithms, constraint programming). Every solution is feasible, scored, and explained per decision.

Can my AI agent call Intellign as a tool?

Yes — that's the point of the Python SDK and MCP server. When your agent needs an optimization answer, it calls Intellign's solve endpoint and receives structured assignments with rationale. pip install intellign.

How long does a solve take?

Typical problems (tens to hundreds of resources) solve in seconds. You watch convergence live, and you get notified in-app if you've navigated away when it finishes.

Can I trust the results?

Every assignment ships with a human-readable rationale and a score breakdown per goal. You can approve, modify or reject each one. If the primary solver is ever unavailable, results are clearly labeled as approximate — never silently degraded.

Optimization · as a service

The math layer
within AI.

Describe the goal in plain English.

Intellign formalises it, runs the solve, and hands back an assignment your team can defend — every decision explained, every constraint honoured.

Launch app Try the live demo

Early access — onboarding design partnerspip install intellignMCP-native

Plain language in.

No solver syntax. No modelling expertise. Operators describe; the engine formalises.

Optimal out.

A solver engine that picks the right method per problem — assignment, scheduling, or routing — and validates every input before it runs.

Explainable.

Every output ships with a human-readable rationale. Auditable by ops, defensible to compliance.

The product

Two ways
to work.

Same engine underneath. Pick the surface that matches the user — a chat for operators moving existing tasks in, a structured playground for engineers wiring the solver by hand.

For operators and enterprises

Move existing optimization tasks into a chat.

Hospital schedulers, logistics ops, public-sector planners — describe the goal, drop in your CSVs, and Intellign builds the problem with you. No solver syntax, no modelling expertise, but every step shows its work: the data it ingested, the columns it inferred, the solver it picked and why.

Active-problem context bar — readiness checks tell you exactly what's missing.
Optimization runs in a sliding canvas: monitor live, expand goals, review assignments without losing your place.
Every assignment row carries a rationale — auditable by ops, defensible to compliance.
Dataset preview and solver pick are inline in the conversation, never behind a modal.

Try smart chat

Intellign AI

Active problemHealthcare deployment124 resources · 48 targetsReady

✓ Data✓ Solver✓ Goals✓ Ready

I've inspected the workspace and found two tables that look like the right inputs. Confirm and I'll structure the dataset.

Accept and continue.

Selected solver: Genetic algorithm · pop 256 · max 2,000 gens

Illustrative interface — explore the real thing in the live demo.

Proven on real problems

Built against a national-scale deployment.

Intellign's pilot case study assigns NYSC healthcare graduates to primary health facilities across Nigeria — matching specializations, honouring capacity, prioritising disease burden. It's the same scenario you can click through in the live demo.

3 turns

from first message to optimization run

~40s

to generate a realistic two-table sample dataset

50/50

resources assigned in the pilot solve, every one explained

13+

file formats ingested — CSV, Excel, Parquet, GeoPackage…

Walk through the case study →

How it works

Five steps from question
to assignment.

The same five steps run whether you're rostering nurses, routing deliveries, or placing teachers. Intellign handles the translation and the solve; you handle the goal and the review.

Describe

Plain English. Tell Intellign the goal, the constraints, and what counts as a good solve.

Ingest

Drop in your CSV, point at an API, or stream from your system. Columns are inferred and explained.

Translate

Goals + constraints get formalised into a structured optimization problem.

Solve

The engine picks the right solver for the problem and runs it — typical solves finish in seconds.

Explain

Every assignment ships with a human-readable rationale. Auditable, defensible, exportable.

Example“Build next week's roster for 24 nurses across 3 wards. Maximise fairness, cap overtime at 8h, honour all leave requests.”→structured roster + rationale,
returned in 4.2s.

Use cases

One engine.
Multiple industries.

The same Intellign call routes deliveries, schedules nurses, places teachers, allocates field officers. Anywhere a human is matching resources to demand by hand or by Excel, the engine fits.

Healthcare

Schedules and routes for care.

Schedule nurses across wards · Assign patients to doctors · Plan home-care visit routes · Optimise ambulance dispatch.

Logistics

Routes that save fuel and hours.

Plan multi-stop delivery routes · Assign drivers to deliveries · Reduce fuel + idle costs · Re-optimise on the fly.

Public sector

Fair allocation at scale.

Assign field workers to regions · Plan rotating staff schedules · Allocate scarce resources fairly.

Education

Timetables that just work.

Build school timetables · Place teachers across subjects · Balance classroom utilisation.

Business ops

Workload balanced.

Assign tasks to employees · Manage team workload · Plan operations efficiently.

For developers

One call from any LLM.

Intellign ships a Python SDK and an MCP server your assistant can call as a tool. Describe the problem, post the resources, receive an assignment.

When a healthcare model needs to schedule nurses, it calls Intellign. When a logistics model needs to route deliveries, it calls Intellign.

See full reference MCP guide

# pip install intellign — call it from any LLM or agent framework.
from intellign import Intellign

client = Intellign()

result = client.solve(
    goal="Minimize total nurse overtime",
    constraints=[
        "Each ward needs ≥ 3 nurses per shift",
        "No nurse exceeds 48h/week",
    ],
    resources=nurses,
    targets=wards,
)

# → 50/50 assigned, every decision explained
result.assignments

Pricing

Three tiers, plus a
developer revenue layer.

Start free, grow into Pro when you're moving real production work, and let your engineering team call the API directly when an LLM needs an optimisation answer.

Free

$0/ month

For individuals exploring the platform.

Capped optimization prompts
Limited dataset size
10 datasets per month
Community support

Start free

Pro

$15/ month

For SMEs and growing teams.

Higher usage limits
Faster processing
Scheduling, routing, assignment
Unlimited dataset creation
Email and priority support

Choose Pro

Enterprise

Custom

For regulated industries.

Unlimited usage
Custom integrations
Dedicated support and SLAs
Advanced compliance and constraints

Talk to sales

API · developers

$0.10/ solve

A revenue layer for AI builders.

Subscription-based API access
MCP and function-calling native
Long-term enterprise contracts

Read API docs

FAQ

Questions, answered.

Assignment, scheduling, matching, allocation and routing problems — who does what, who goes where, what gets which share. Staff rostering, field-worker deployment, teacher placement, delivery routing, budget allocation.

Get started

Every organization needs
optimization. We'd like you in
when they start asking AI for it.

Launch the app, describe a problem, and run your first solve — three chat turns from question to an explained assignment.

Launch app Book a demo

The math layerwithin AI.

Two waysto work.