Siddharth Karandikar · Strategy & engineering
Most problems are only hard at the wrong resolution.
I'm a final-year Electronics & Telecommunication engineer who does strategy work. At Godrej Properties that meant a 553-component cost master, and ₹178 crore of a ₹183 crore savings proposal approved. In January 2027 I join Searce as an AI Solutions Consultant.
Selected work
I find money by breaking things down further than anyone had.
A cost base split into 553 components, a market read from seven years of customs records, an overseas entry taken in stages. Each time, the answer appeared at a finer resolution than the one people were arguing at.
97% of the savings proposal was approved
Hover, tap or tab to a bar for its value.
Source: Godrej Properties, 2026 — approved savings against the proposal
- ₹178 crore of a ₹183 crore savings proposal was approved, because the cost base had been broken down far enough to argue with line by line. MEP cost optimisation · Strategy Intern ₹178 Cr Savings approved, of ₹183 Cr proposed
- ₹1,700 crore of import records contained ₹30 crore of demand nobody had called. Demand discovery in EDM machines · Competitive Strategy Intern ₹30 Cr+ Untapped demand identified
- Enter a foreign market as somebody's representative first, and as a principal later. Live Staged entry into an overseas market · Strategy Intern Staged Representation first, capital commitment later
- Rural banking expansion that assumes the internet is not there. CEO Challenge · Winner, three-person team 1st / 25+ Against MBA and undergraduate teams
The technical edge
I can tell which parts of a pitch are physics.
More of the problems strategy teams are hired for now have technology inside them: an operator's network spend, a bank's payment rails, an AI rollout. I have built in those systems, so I can read the technical half of the business case instead of taking it on trust.
- The question is not whether a model predicts congestion in simulation. It is whether that model survives contact with real network data. In progress Telecom operators · Network equipment vendors 250 ms Telemetry granularity, 50 devices over four cells
- A framework nobody could reproduce now runs 64 times cheaper, with three errors in the original corrected. Telecom · Automotive · Defence and drones 64× Reduction in dominant computational cost
- A commodity microcontroller holds clinical timing if the buffering is designed for it, and false alarms are a design target, not a side effect. Medical devices · Healthcare 3,906 µs Interrupt-driven sampling interval, jitter-free
- You can validate an engine control unit without owning an engine. Automotive · EV and component suppliers Bench Full validation with no vehicle required
- Payments survive peak load only if the network slice is reallocated while the surge is happening. Banking and payments · Telecom operators 500,000+ Concurrent transactions simulated
One line, one turn
Two threads since 2024, both pointing at the same job.
In 2024 I spent six months building a startup and found that the part I wanted was reading a business from the outside. The engineering did not stop; the hardest radio work came after. A second thread started beside it.
- Before the turn MindSpark Sponsorship for a ₹30 lakh festival. The first commercial work. Aug – Oct 2023.
- Before the turn Police workstation User research for traffic police. The constraint was a place to work, not equipment. Jan – May 2024.
- Before the turn B-ABLE programme A hundred students taught, once the permissions were secured. 2024.
- The turn TimeWarp Six months building a startup and pitching it to angel investors. The turn toward strategy. 2024.
- Commercial CEO Challenge Rural banking built on the assumption of no connectivity. First place. Jan 2025.
- Engineering COEP 5G Lab ₹40 lakh of government funding brought onto campus. Jan – Apr 2025.
- Engineering Banking over 5G slices Half a million payments pushed through network slices to find the failure point. 2025.
- Engineering Automotive ECU A control unit validated on a bench, with no vehicle attached. Feb – May 2025.
- Commercial Sparkonix ₹30 Cr of untapped demand found in seven years of import data. May – Jul 2025.
- Engineering EEG seizure detection Clinical sampling timing held on a commodity microcontroller. Jul – Dec 2025.
- Commercial Pune Startup Fest Investor commitments up 54%, on the strength of a screening funnel. Sep 2025 – Jan 2026.
- Engineering OTFS-ISAC research Reproduced a framework published without code, and corrected three of its results. 2026.
- Engineering O-RAN congestion prediction A model trained in simulation and tested on real network data. 2026, ongoing.
- Commercial Godrej Properties ₹178 Cr of a ₹183 Cr savings proposal approved. May – Jul 2026.
- Commercial Overseas market entry A staged foreign market entry, currently live. 2026.
- Next · Jan 2027 AI Solutions Consultant, Searce Where both threads meet.