In one of the science parks somewhere in Taipei, the R&D building is still lit up in the dead of night. The blue light from the screens casts on the tired faces of the engineers as they pour over thousands of rows in Excel spreadsheets, opening each PDF datasheet, copying values into cells one by one, and checking that MCUs, sensors, and other components will work together. It’s all manual. In the electronics manufacturing industry, such scenarios are not uncommon: a single small mistake can often delay an entire production line, resulting in losses of hundreds of thousands of dollars.
“Engineers often have to spend over 100 hours going through as many as 2,000 pages of datasheets, just to verify that the hardware components are compatible,” said Niyam AI co-founder and CEO Samarth Shyam. “In this industry, every day of delay can cost hundreds of thousands of dollars, and for brands, a month’s delay in launching a product can result in a loss of 20% to 30% of their lifecycle profits. Across a portfolio, those slips compound to millions, if not billions.”
The electronics manufacturing industry is the backbone of global innovation, yet it has long been hampered by inefficient design tools. The first-pass success rate of prototyping is less than 1%, and engineers not only have to check the compatibility of components one by one but also deal with downstream issues such as parts obsolescence and supply chain instability.
An even bigger problem is that electronic design involves multiple teams but lacks a unified platform to integrate the entire process, resulting in many problems only discovered after the prototype has been completed. Existing tools lack real-time feedback and proactive analysis, leaving companies often forced to “put out fires” only after mistakes have been made.
And this is the very pain point that Niyam AI is trying to solve.
How can two atypical hardware professionals disrupt the electronics manufacturing industry?
The two founders of Niyam AI are not typical “hardware folks,” but they have recognized the industry’s long-standing pain points.
CEO Samarth Shyam is a serial entrepreneur and angel investor with a track record of shipping and exits and plenty of battle scars; his previous company sold tokenized loyalty platforms to enterprises and was acquired. CTO Agrim Singh was Citibank’s first Hacker in Residence and built market analytics that surfaced price-moving events minutes before Bloomberg or Reuters.
The two met in the Entrepreneur First program in Singapore and have since worked on a number of projects, but the real turning point came after watching a founder friend nearly lose his startup because of a single wrong MCU choice. It was an upstream failure. Kickstarter is a graveyard of similar stories where tiny part data mistakes snowball into missed launches.They found that hardware prototypes often required multiple revisions due to component compatibility or specification errors, and engineers lacked tools that could “proactively” identify problems and were forced to make revisions only after the errors had occurred. This was a pure data (ETL) problem and they knew it!
This led them to the decision to found Niyam AI, which catches component risks early by surfacing compliance, sourcing, and design errors before they turn into costly surprises. The duo combines commercial GTM firepower with deep AI execution, the mix needed to win a fast moving, niche technical market.
“The AI spellcheck for the World of Hardware” makes design ten times faster! What does Niyam AI do?
“Niyam is the AI spellcheck for the world of hardware.” Shyam explained that engineers, who used to spend hundreds of hours comparing data and copying specs out of PDFs, now have Niyam sitting inside ECAD and PLM. The agent pulls design files, BOMs, and linked datasheets automatically, then parses and checks them in minutes.
Shyam further explained that at the beginning of the process, the agent extracts and normalizes specs from each datasheet and design file, builds structured part data, and runs instant compatibility checks. It then looks for risks like insufficient inventory, long lead times, and expiring certifications.
All test results are consolidated into a single dashboard, where engineers can simply click on the suggested options and have an updated bill of materials automatically generated and imported back into the existing Product Lifecycle Management (PLM) system, with little or no change to their work habits.
On the supply-chain side, Niyam has significantly reduced the cost of manual cross-checking. On the supply chain side, the agent removes most manual cross checking. In the past, teams entered part numbers across thousands of suppliers, verified inventory and certifications, and picked replacements by hand.
Now, Niyam’s agentic AI will start from the bill of materials and proactively provide “contextualized recommendations”not only telling you “what parts are available” but also filtering out solutions that “really work,” based on criteria. This proactive, real-time feedback makes design over ten times faster, revolutionizing the reactive, time-consuming model of the past.
How can Niyam AI enter the EDA market dominated by giants?
Over the past 30 years, the EDA market has been monopolized by giants such as Synopsys, Cadence, and Siemens, with tools primarily focused on design. Although they do have verification features, these are limited to basic checks. As hardware design becomes more complex, traditional tools struggle to perform cross-component analysis at an early stage and lack data integration with the supply chain. EDA vendors make money from simulation seats, and PLM vendors make money from change-order churn, so proactive upstream verification has not been a priority.
In the wave of AI, many startups such as Flux, Celus, and JITX have tried to enter the market, but most focus on design efficiency but they leave the core problem untouched: part data still moves by hand and compatibility gets verified late.
Niyam, on the other hand, targets the upstream failure. Their agent sits inside the ECAD and PLM, pulls schematics, BOMs, and linked datasheets automatically, then runs continuous risk checks for compatibility, lifecycle, sourcing, and compliance. No uploads, no one-off scans, just background guardrails that surface issues early and propose fixes. Compressing what would otherwise take hundreds of hours into minutes.
This positioning has allowed Niyam to quickly establish an advantage: on one hand, it has accumulated expertise in component lifecycle, supply-chain dynamics, and design rules; on the other, it has secured validation from major EMS’s and ODMs within a year of its founding, continuously strengthening its data, processes, and trust in real-world scenarios to build a high barrier to entry.
“In enterprise SaaS, one of the hardest things is figuring out who the real users are, and where in the value chain you fit in”
In November 2024, the team rebuilt the product to run as an agent inside the customer’s environment.
Upload-based workflows hit a wall in practice. “A BOM or schematic can have over 300 permutations in the lifespan of a product. No one is going to upload a new file each time. If you rely on uploads, most risks stay invisible,” Shyam said. Rather than trying to change habits that are decades old, the team made the workflow agentic. “You meet engineers where they work. The agent watches ECAD and PLM for changes, pulls the right datasheets, runs the checks, and proposes fixes automatically,” he added.
Adoption improved dramatically once those extra steps disappeared. Early versions asked teams to upload files to a dashboard, which worked once in a while, not every day. The new flow plugs directly into ECAD and PLM and serves results in place. The system still supports Excel and CSV for teams that need them, but the default is no manual uploads.
The team also learned who the real day-to-day users are. “At first we thought the daily users would be design engineers,” said Shyam. “In practice, librarians, component engineering, PLM, sourcing, and compliance live in these tools. They care most about accuracy and traceability. If it takes hours, they will still choose correctness.” Time pressure sits with downstream brands and program owners. “For an international brand like Apple, a delay can miss a market window. A month can cut lifecycle profit by 20 to 30 percent,” Shyam said.
Those lessons shaped the product. Meet the bar for accuracy that front line teams demand, run inside their systems so adoption is painless, and deliver time savings leadership can feel. “That is how Niyam becomes part of the supply chain instead of another dashboard people open once and forget,” Shyam said.
Niyam: Creating a “New Order” for the Manufacturing Industry
After a run of pilots and validations, Niyam drew interest from major EMS and ODM partners and was selected for the Wistron Accelerator in 2025. The program let the team test and iterate in real factory conditions and signaled meaningful validation from industry leaders. That selection led to a full sitewide pilot with Wistron, one of the largest manufacturers in the world, and pilots in the pipeline with three of the ten biggest manufacturers globally.
Rather than just catching errors, Niyam reshapes the engineer’s workflow. The agent sits inside ECAD and PLM, surfaces risks early, proposes fixes, and writes clean revisions back into the system of record. “In this industry, even a two to three percent improvement in efficiency can translate into billions of dollars in revenue,” said Shyam.
For Shyam, this is the starting line. The vision is clear: in five to seven years, Niyam will be the agentic AI backbone of the two trillion dollar electronics supply chain. “Every tier 1 EMS, ODM, and OEM runs designs through us, from datasheets and BOMs to alternates, certifications, even placement. What used to be manual, siloed, and error prone becomes self healing and automated. Hardware finally iterates as fast as software, and if you are building electronics anywhere in the world, Niyam is quietly in the background making sure nothing slips,” said Shyam.
The name Niyam comes from Sanskrit for rules and a way of life, and that is the aim: turn the rules that keep hardware safe into everyday practice inside ECAD and PLM so discipline becomes the default across the supply chain.
