Skip to content
TiMiNa
TiMiNaAgentic AI Workflows for FMCG

The Agentic FMCG Enterprise, built with the people who run it.

We build teams of AI agents that sense demand, replenish stores and guard margin. Your people keep every decision that matters.

01The problem

Thousands of decisions a day, most of them made in spreadsheets.

What to order, where scarce stock goes, which price holds. Your best planners spend their days on routine, not judgment.

Routine decisionNeeds a person’s judgment
02Our approach

Agents take the routine. People keep the judgment.

We rebuild one workflow at a time, wherever the routine lives in your value chain.

Illustration of a factory, distribution centre and stores connected by delivery routes
01 / 3L2 · Act on approval
Allocation agent

Shares scarce DC stock by a declared, fair rule.

03How we keep it safe

Agents prepare. People decide.

Every system we build is drawn in three shapes, so anyone can see who prepared a decision and who owns it.

Data is reconciled, never guessed.When a figure is missing, the system says so.
Agents work inside hard limits.Every run leaves a record you can trace and reproduce.
A named person approves every action.Nothing is ordered, priced or changed without sign-off.

Autonomy, set per decision

It starts low and rises only on measured evidence.
L0Observe
L1Recommend
L2Act on approval
L3Act within guardrails
L4Autonomous: not deployed for accountable decisions
04Proof

Built and tested in our first year.

Three systems, each shown by the number that matters most. Client names are left out by agreement.

0corrections to our map of their process
Grocery retail chain

Store replenishment from the DC

Built and tested
11stages of accuracy testing before any claim
FMCG distributor

Sales intelligence and forecast accuracy

On 5+ years of real data
14guardrails on every price proposal
Promotional products supplier, Europe

Pricing intelligence

Designed and planned
05How we start

One workflow first. Proven, then wider.

Every step ends with something your team sees and decides on.

  1. 01

    Map

    We map how the workflow runs today, with the people who run it.

    You confirm the map.
  2. 02

    Agree the rules

    Every rule and decision right is written in plain language.

    You sign off each one.
  3. 03

    Prove

    We build on sample data shaped like yours and measure the results.

    You review real outputs.
  4. 04

    Go live

    Agents recommend. Autonomy rises only on evidence.

    You approve every action.
06The team

Senior people, accountable for your work.

Commercial FMCG depth, technology leadership at scale and hands-on agentic engineering.

Misagh Akhondzad

Co-founder and CEO

Leads client work: turning how an operation really runs into rules agents can follow.

  • 20+ years of commercial and revenue leadership
  • Deep FMCG commercial and go-to-market experience
  • Has run a B2B business with full P&L

Norman Noble

Co-founder, technology

Leads how we build: architecture, engineering standards and delivery.

  • CTO for global consumer and financial services businesses
  • Builds and leads high-performing engineering teams
  • Co-founder of agentic AI and B2B SaaS ventures

Nasir Shadravan

Senior solution architect

Designs how agents connect to your data, your ERP and each other.

  • 15+ years of software engineering
  • Builds and runs AI products with paying customers
  • Specialist in LLM applications and multi-agent workflows

Alongside us, senior agentic workflow developers build, test and run every system.

Data informsAgents preparePeople decide

The routine goes to agents. The judgment stays with your people.

Tell us which workflow takes the most routine effort. We will show you how we would approach it, and what you would see first.

Misagh Akhondzad · misagh@timina.ai