HI, I AM JITIN
HI, I AM JITIN
Your AI is running. The dashboards are full. Six months later, nobody’s changed a decision because of it. That’s not an AI problem. That’s a Logic Leak and it’s fixable.
Your AI is running. The dashboards are full.
Six months later, nobody’s changed a decision because of it.
That’s not an AI problem. That’s a Logic Leak — and it’s fixable.
I’m an independent AI consultant. 15 years. $90M+ in portfolios. I find the leak.
The same diagnostic process that recovered $35M for a global FMCG — correcting the decision that had been making the wrong call for 18 months — works across operations. Because the pattern is always structural, not technical.
The case that made the pattern visible
An automotive company was losing deals. Their sales cycle conversion sat at 4% — which everyone in the room had accepted as the natural rate for their category. The data science team had built a binary prediction model: buy or no-buy. It was correct about 60% of the time. No one could work out why it wasn’t helping.
The frame was wrong. A sales cycle is not a binary event. It is a sequence with friction points — exactly like a manufacturing process under load. When I stopped asking “will this person buy?” and started asking “at which point does this interaction stall, and why?” — the problem became solvable.
I mapped the sales cycle as a time-to-failure model, the same method used to predict when a machine on a production line will break down. The model identified, for each lead, the precise stage where conversion would stall and what intervention would prevent it.
Conversion moved from 4% to 11%. That is a 2.7x lift in revenue. Without a new platform, a new team, or a new data source.
The data was there the whole time. The frame was the problem.
4% → 11%
Conversion rate improvement
2.7x
Revenue lift
Zero new infrastructure
Built on existing data
The pattern I keep seeing
“Most teams hire people who only know their industry. So they keep solving 2026 problems with techniques from 2010 — because nobody told them to look sideways.”
Every organisation I have worked with arrived carrying one of three problems. Sometimes all three at once.
Confusion. The gap between what AI is marketed to do and what it can actually do in a specific operational context is enormous. Leaders are being asked to make multi-million-dollar platform decisions with no framework for evaluating them. The result is not bad technology — it is bad procurement.
Fear of the wrong thing. The question executives ask privately is not “will AI replace me?” It is “will I make the wrong call and spend three years recovering from it?” That fear is rational. It is also paralysing. The way out is not more information about AI. It is a method for making a defensible decision.
No number. Organisations run pilots without quantifying what success looks like. When the board asks for ROI, the answer is a range of assumptions dressed as a forecast. That is not defensible. It is the fastest way to lose board confidence on AI investment.
These are not technology problems. They are structural ones. And they have structural solutions.
What I believe
The most valuable AI work is rarely the most visible. Predictive logic, optimisation, and structured decision systems quietly move more P&L than any LLM demonstration. I build for the quiet 70%, not the highlight reel.
I am not a generative AI specialist. I am not a machine learning researcher. I am an AI strategist — which means my job is to identify the right tool for a specific operational constraint and build the architecture that connects it to a business outcome.
The right tool is often not the newest one. In the automotive quality control engagement, the answer came from cardiology, not industrial AI. In the FMCG forecasting case, the fix was a signal processing decision, not a model upgrade.
I am an AI resident, not a tourist. I still work in production environments. I read the code. I understand what the engineering team is actually building and I can translate it into language the CFO will act on. That is the combination most organisations are missing — not more AI capability, but a person who can sit in both rooms and not lose anything in translation.
Most companies have been running with a Logic Leak for 12–18 months before anyone names it. A 30-minute Clarity Call tells you whether that’s you — at no cost, no pitch, no obligation.
“He doesn’t deliver a deck and disappear. He stays until the logic is in the system and the team can run it. That’s rare.”
— Nikhil Jain, Head of Tech Accelerator & AI Innovation, L’Oréal
“15% profit margin growth in six months. We’d spent 18 months on dashboards that told us what happened. Jitin shifted us to systems that change what happens next.”
— VP Operations, Zespri (NDA — name withheld)
The method
Three steps. Every engagement uses them. In sequence.
1 — Test the use case before building it Before any build decision, three questions are answered: Is this the right problem given the operational constraint? Does the data actually exist to support it? Is the timing right given everything else in motion? Most organisations skip this step. That is where the wrong vendor contract begins.
2 — Specify it in business terms, not technical ones Every use case worth building can be written in four terms: what outcome is being optimised, what business logic connects input to decision, what operational constraints the solution must respect, and what data actually exists. If it cannot be written this way, it is a hypothesis — not a brief.
3 — Put a number on it before spending one A structured model converts the use case into a P&L figure before a line of code is written. Not a forecast. A specific, defensible number built from the actual constraints of the business — in language a CFO can interrogate without a data scientist in the room. This is what ends the “we’ll see what the pilot shows” conversation.
The background
I trained as a mechanical engineer — which is where the cross-domain instinct comes from. Fifteen years ago I moved into AI strategy because operational problems kept arriving at my desk that the standard toolbox couldn’t solve.
The automotive case used biomedical algorithms. The FMCG case used control systems theory. A telecom case I worked on in the same period used the same survival models that predict equipment failure in power plants. The right answer rarely lives where you expect it.
Since then: $90M+ in AI programmes managed across manufacturing, FMCG, logistics, telecom, and automotive. Clients include Fortune 500 companies across Europe, Asia-Pacific, and the Middle East.
I write about AI strategy and the decisions behind it on the blog and in the Weekly AI Decision Brief — where 15 years of deployment experience is the source, not secondary research.
I am based in Bangalore and work entirely remotely with organisations globally — primarily Australia, New Zealand, UK, Europe, and the US. Every engagement runs online. No site visits required.
Connect on LinkedIn →
Who I work with
This works best for:
COOs, VP Operations, CFOs, CHROs, and founder-MDs who carry P&L responsibility and are facing one of three situations: they need to make an AI investment decision and don’t have a framework for evaluating it; they have already invested and the results are not showing up in the P&L; or they are managing an active AI programme and need an architect in the room — not another vendor.
Company size: 50 to 5,000 employees. Operations-heavy industries. The work is not sector-specific — the frameworks travel.
This is not for:
Organizations looking for a vendor recommendation or a platform shortlist. Technology teams looking for a data science lead. Leaders who have already signed the vendor contract and need someone to make it work. I can refer you elsewhere for all three.
“I don’t start with tools. I start with the business logic, the constraints, and the question — then identify the simplest system that answers it.”
The Weekly AI Decision Brief
One sharp insight on making AI work for your business — every week. Written for operations leaders, not technology teams.
Work with me
AI Profit OS
3-day sprint. Ranked use case list, ROI formula, board memo.
From $1,500 per seat
AI Strategy Audit
2–3 week diagnostic. Logic Leak identified, 12-month roadmap, vendor guidance.
From $2,400
Weekly AI Decision Brief
One sharp insight on making AI work for your business — every week. Written for operations leaders, not technology teams.
Free Forever