Praveen Jonnala, SVP and CIO of Vistance Networks, argues in a CIO.com opinion piece that leading AI labs and cloud providers are misdirecting their enterprise AI investments by sending forward-deployed engineers to customers. He writes that the approach solves the problem of getting AI pilots into production but does not address how AI diffuses through the wider economy.

Jonnala cites an MIT finding that about 95% of enterprise AI pilots produced no measurable profit impact despite tens of billions of dollars spent. In response, OpenAI launched its OpenAI Deployment Company in May with $4 billion of initial investment from 19 partners, starting with 150 forward-deployed engineers (FDEs) and deployment specialists. Anthropic partnered its engineering capability into an enterprise AI services venture with Blackstone, Hellman & Friedman, and Goldman Sachs, reportedly valued above $1.5 billion. Microsoft launched Frontier Company in July with $2.5 billion and approximately 6,000 embedded engineers and industry experts. AWS committed $1 billion to a forward-deployed engineering unit. FDE job postings grew roughly 800% in nine months, with principal-level compensation past $1.2 million.

Jonnala says the forward-deployed function sits on the commercial side of these companies and measures pilots converted to production within the quarter. He argues that such engagements optimize for accuracy or local workflow results rather than cash outcomes, which depend on decisions spanning several functions and appear in cash only two quarters later. He describes a demand-forecasting agent whose accuracy gains changed nothing financially until operational decisions such as safety stock and supplier commitments were altered.

He also raises structural concerns. Anthropic's job listings ask engineers to codify repeatable deployment patterns and feed insights back to product teams, which he presents as a sign that the learning flows toward the lab. Constellation Research warned this year that FDEs are often used as a crutch for product immaturity. Dice reported in August that the engineer shortage is holding back enterprise AI returns, with the scarcest talent concentrated in Fortune 50 accounts while mid-market companies wait. Jonnala's test for any engagement is whether the customer becomes better at finding its next problem after the outside engineers leave.

Jonnala describes his own company's inventory fix, which came from redesigning its planning architecture with operators who knew the supply chain, rather than pointing an off-the-shelf frontier model at raw ERP tables. Inventory fell and stayed down, and no outside team was needed to repeat the work. He concludes that a country's AI advantage will depend on millions of operators applying AI to operational decisions, and that companies should send business operators to learn AI rather than AI experts to learn business. He calls for backward-deployed domain experts, a role he says no vendor is selling.

Praveen Jonnala, SVP and CIO of Vistance Networks, argues in a CIO.com opinion piece that leading AI labs and cloud providers are misdirecting their enterprise AI investments by sending forward-deployed engineers to customers. He writes that the approach solves the problem of getting AI pilots into production but does not address how AI diffuses through the wider economy.

Jonnala cites an MIT finding that about 95% of enterprise AI pilots produced no measurable profit impact despite tens of billions of dollars spent. In response, OpenAI launched its OpenAI Deployment Company in May with $4 billion of initial investment from 19 partners, starting with 150 forward-deployed engineers (FDEs) and deployment specialists. Anthropic partnered its engineering capability into an enterprise AI services venture with Blackstone, Hellman & Friedman, and Goldman Sachs, reportedly valued above $1.5 billion. Microsoft launched Frontier Company in July with $2.5 billion and approximately 6,000 embedded engineers and industry experts. AWS committed $1 billion to a forward-deployed engineering unit. FDE job postings grew roughly 800% in nine months, with principal-level compensation past $1.2 million.

Jonnala says the forward-deployed function sits on the commercial side of these companies and measures pilots converted to production within the quarter. He argues that such engagements optimize for accuracy or local workflow results rather than cash outcomes, which depend on decisions spanning several functions and appear in cash only two quarters later. He describes a demand-forecasting agent whose accuracy gains changed nothing financially until operational decisions such as safety stock and supplier commitments were altered.

He also raises structural concerns. Anthropic's job listings ask engineers to codify repeatable deployment patterns and feed insights back to product teams, which he presents as a sign that the learning flows toward the lab. Constellation Research warned this year that FDEs are often used as a crutch for product immaturity. Dice reported in August that the engineer shortage is holding back enterprise AI returns, with the scarcest talent concentrated in Fortune 50 accounts while mid-market companies wait. Jonnala's test for any engagement is whether the customer becomes better at finding its next problem after the outside engineers leave.

Jonnala describes his own company's inventory fix, which came from redesigning its planning architecture with operators who knew the supply chain, rather than pointing an off-the-shelf frontier model at raw ERP tables. Inventory fell and stayed down, and no outside team was needed to repeat the work. He concludes that a country's AI advantage will depend on millions of operators applying AI to operational decisions, and that companies should send business operators to learn AI rather than AI experts to learn business. He calls for backward-deployed domain experts, a role he says no vendor is selling.