BRIEF26 | Shailesh Dhuri of Decimal Point Analytics: Build relationships that provide context algorithms can’t access

This Year-Opener under BRIEF26 has been authored by Shailesh Dhuri, Co-founder and CEO, Decimal Point Analytics, exclusively for Mediabrief.com. Insights and inspiration for tomorrow.
2025 was the year we stopped asking “Will AI replace knowledge workers?” and started asking a harder question: “Which knowledge workers are worth augmenting?”
The lesson from 2025 wasn’t about technology. It was about us.

What 2025 Taught

The most valuable insight came from implementing Statistical Process Control across our analytics operations. SPC forces you to distinguish between signal and noise, between systemic problems worth fixing and random variation you should leave alone. The best performers weren’t those who followed procedures most diligently. They were the ones who understood why the procedures existed, and knew when to override them.
This distinction matters enormously in an AI age. Large language models are exceptional procedure-followers. They’re terrible at knowing when procedures don’t apply. The professionals who will thrive aren’t competing with AI on speed or recall. They’re developing the judgment AI lacks.
Another uncomfortable truth: much conventional wisdom has expired. “Build proprietary databases” made sense when information was scarce. Today, information is abundant; interpretation is scarce. “Hire for domain expertise” assumed expertise was stable. Now, the half-life of technical knowledge is shorter than the tenure of most employees.

What 2026 demands

Four shifts worth watching: First, the low-altitude economy is arriving. Through my involvement with Empyreal Galaxy, an airships and aerostats venture, I’ve seen how dramatically the economics of persistent aerial platforms are improving. Surveillance, communications, cargo delivery, atmospheric research, the airspace below 20,000 feet is becoming infrastructure. India, with its logistics challenges and growing drone ecosystem, should be paying closer attention.
Second, energy storage is crossing thresholds that change everything. Sodium-ion batteries are making grid storage economical. Solid-state batteries will reshape electric vehicles within this decade. Solar costs continue their relentless decline. The firms still planning around legacy energy assumptions are building on sand.
Third, robotics is following AI’s trajectory, slowly, then suddenly. The humanoid robots demonstrated in 2025 aren’t ready for deployment, but the trajectory is clear. Combined with advances in computer vision and manipulation, physical automation will transform manufacturing and logistics faster than most business plans assume.
Fourth, own the data layer. India’s financial services sector is building digital rails at unprecedented speed. The firms that own the plumbing—the registrars, the reconciliation systems, the audit trails, will shape the industry for decades. Everyone wants to be the interface. The smarter play is being the infrastructure.

To the young professionals

You’re entering industries being reshaped by tools that can do much of what juniors traditionally did. This is not a threat. It’s a liberation.
The mundane work that consumed your predecessors’ early careers, data cleaning, template formatting, routine calculations, will increasingly be handled by machines. Your job is to develop the judgment those machines lack.
Ask questions that don’t fit the template. Notice what the data isn’t saying. Build relationships that provide context algorithms can’t access.
And please: stop optimizing for looking busy. The professionals who will matter in 2026 are those who can sit with ambiguity, resist premature conclusions, and tell clients uncomfortable truths. These are human capabilities. They’re not being automated away.
As I approach my 56th year, I’m more convinced than ever that the future belongs to those who can hold two ideas simultaneously: deep respect for quantitative rigor, and deep skepticism that any model captures reality.
The centaur that wins isn’t the one with the fastest AI. It’s the one with the most calibrated rider.