Data. Intelligence. Results.

Three words, in that order — and it's a process, not a slogan. Take in the messy reality of an operation, find what actually matters inside it, and hand back something you can act on Monday. The domain changes: a warehouse floor, a federal feed, a pile of spreadsheets that disagree with each other. The process doesn't.

How we work

One process, three stages

Every engagement runs the same way, whatever the industry. It's the reason the work is checkable — you can see which stage a number came from.

Data — take in the messy reality

Your WMS export, a federal solicitation feed, four spreadsheets nobody trusts. We start with what you already have, however unusable it looks, and consolidate, clean, and validate it into something that can actually be reasoned over. Nothing gets invented to fill a gap — missing is recorded as missing.

Intelligence — find what actually matters

The analysis layer. Where there is a right answer, it gets computed rather than guessed. Where it takes judgment, it is labelled as judgment with the reasoning attached. This is the stage that turns a dataset into a dollar figure, a shortlist, or a ranked set of moves.

Results — something you can act on Monday

A deliverable a person can use without a manual: the twenty changes worth making, the three opportunities worth pursuing, the report that now writes itself every week. Every number traces back to your data, so any of it can be checked.

What we do

The same three stages, pointed at different problems

These are applications of the process, not separate products. If your problem isn't listed, it's probably still the same three stages.

Warehouse & logistics operations

Pick-labor cost, travel paths, slotting decay. We model your real floor and compute the provably shortest pick or travel route — exact math, not an AI guess — then hand back a prioritized list of changes worth making.

Data operations & automation

Scattered files and exports merged into one reliable master dataset, then the repetitive analysis and reporting set to run on their own — fixed-scope, reviewed with you before anything is final.

Federal contracting

For federal buyers: a registered small business you can put a defined piece of data or analysis work to. Active in SAM.gov, fixed scope, fixed price, and the measurement agreed before anything starts.

Operational problems that don’t have a name yet

The recurring decision nobody has time to make properly, the number the business runs on that nobody can trace. Bring the problem in plain words; we’ll tell you honestly whether the process fits it.

Built and running

Pipelines, not slideware

Systems that exist and run today. Each is labelled with what it actually is — in production, or in demonstration. No client names, no borrowed logos.

Built · paused

Federal opportunity sweep

An automated pipeline that pulls federal solicitation and grant feeds, filters them against a capability profile, and produces a qualified shortlist with a bid / no-bid read. It flags what it isn’t sure about rather than guessing. It ran nightly and unattended; the schedule is paused while attention is on direct client work, and it runs on demand.

Data → scraped feeds · Intelligence → fit scoring · Results → a morning shortlist

In demonstration

Pick-path optimization for a warehouse floor

Built from direct experience running pick operations: it models a real floor’s layout and racking, then computes the provably shortest pick sequence — checked against brute force, not estimated. It quantifies the travel waste in the current sequence in dollars.

Data → order & slotting history · Intelligence → exact route math · Results → a ranked moves list

In production

The system that runs this firm

99 Strategic runs on its own tooling: a multi-agent operations system with an append-only decision ledger, refusal logging when an agent lacks grounds to answer, and a human approval gate on anything that ships, sends, or spends. We build the automation we sell.

Data → operational ledgers · Intelligence → agent analysis · Results → work that ships

What makes the output trustworthy

The same rules apply at every stage — they're the reason a number from us can be checked instead of taken on faith.

Exact where it must be exact

Anything with a right answer is computed, not guessed. Deterministic beats plausible.

Honest when it does not know

Absent facts are flagged, never filled with a confident guess. “I don’t know” is a real answer.

Everything traced

Numbers carry their source and every claim can be checked back to the data behind it.

A human signs off

Nothing ships, sends, or spends without a person approving it. Automation assists; it does not decide.

Fixed-scope, reviewed with you

A clear quote up front and work reviewed with you before anything is final — no open-ended meters.

Your tools, your data stays yours

Built around the systems you already use; confidential handling and working copies deleted on request.

Who's behind this

One operator who builds the tools

99 Strategic is a founder-led practice run by Tuan Le — a career spent running operations and working warehouse and logistics floors, now building the data and automation systems himself. You work directly with the person doing the work.

Tell us the problem in plain words

The operation that never quite runs smooth, the messy dataset, the decision nobody has time to make properly. If the process fits, we'll tell you how. If it doesn't, we'll tell you that instead.