Three products · One mission

AI for the industries that build the world.

Thuniv Industries makes AI products for engineering, construction and B2B sales teams. They sit on the systems and data you already have, cut out repetitive work and help people make faster decisions.

Explore products ↓ Request early access
PLMC
PLM, BOM and cost control for construction
→
Jarvis
The decision layer that runs your shop floor
→
ThadamIQ
Lets small manufacturers sell direct, no distributor
→
The product suite

Three products that work together.

Each one is an agent grounded in your own data — it can act inside your systems, not just answer questions, and it always asks before anything irreversible happens.

01 — Construction PLM Working MVP

PLMC

Plan a building the way engineers plan a product.

Construction projects are still estimated in spreadsheets that fall out of date by the next revision. PLMC brings manufacturing-style PLM to construction: a live structure of towers, floors, units and assemblies, a bill of materials on every item, cost rolled up to the project total — and an AI agent sitting on top of all of it, grounded in your real numbers, that can act on them when you let it.

  • An agent that acts, not just answers — ask it about costs, BOMs, suppliers or quotes in English, हिन्दी or தமிழ். Flip it from read-only to changes-allowed and it can update the plan itself — it still previews every irreversible step and waits for your yes
  • Live cost roll-up — quantity × rate, plus wastage, labour, equipment and transport, totalled up the tree and split by trade
  • RFQs straight from the BOM — the agent can raise one and invite suppliers itself; they quote through a no-login portal for you to compare and award
  • Documents on every node — drawings, photos and spec sheets filed next to the scope they describe
See PLMC on your project →
02 — Decision Layer In design

Jarvis

Say it once. Every machine on the floor hears it.

Jarvis sits between a human instruction — typed or spoken — and the equipment that actually does the work. An intent analyser reads what you meant, Jarvis reasons over two knowledge bases (general domain knowledge, plus what it knows about your own plant) to make a context-aware call, then dispatches coordinated actions to whichever layers need to move: machining centres, PLCs, cameras and LIDAR, robotic arms and drones — all from one decision layer, not four separate systems.

  • One instruction, every layer — machining (CNC, 3D printers), PLC (start/stop/slow a line), visual (camera/LIDAR) and kinematic (robotic arms, drones) all coordinate through the same core
  • Two knowledge bases, one decision — general domain knowledge plus knowledge specific to your industry and plant combine into a context-aware call, not a generic one
  • Closes the loop — every layer reports status back — machine state, position, load, contact — so Jarvis confirms an action actually happened before moving on
Talk to us about Jarvis →
03 — Sales & Growth Live · Revenue

ThadamIQ

Every lead leaves a trace. ThadamIQ follows it to a close.

3paying licences
₹15kMRR

Thadam (தடம்) means trace. ThadamIQ is B2B sales AI built for small and medium manufacturers who want to sell direct instead of going through a distributor. It sources leads, reads each one's own website before writing to them, and sends a personalised first email — not a mail-merge — in 13 clicks. It's simple enough for one person to run with almost no training, and it follows up and reads replies on its own from there.

  • Built for manufacturers, not distributors — sell straight to buyers instead of routing everything through a distributor network, with no separate sales team to hire
  • 13 clicks, start to sent — source a lead, get it researched, and send the first personalised email in 13 clicks. Easy enough to train someone on in an afternoon
  • Runs the loop on its own after that — qualifies, follows up across email and WhatsApp, and reads replies automatically, with a kill switch, daily caps and quiet hours in front of every send
  • GST-ready quotes & invoices — correct CGST/SGST or IGST split generated automatically, exported as PDF or Word
See ThadamIQ in action →
How we build

Built by engineers, not just for them.

We've worked inside OEMs and global engineering centres. We know how slow change can be on a factory floor, and we build for how these teams actually work.

01

Start from the problem

Every product starts with a real bottleneck we've seen on the floor, not with a technology looking for a use.

02

Grounded, not guessed

In engineering a wrong answer costs a lot. Every output can be traced back to its source.

03

Fits your systems

We connect to what you already run — PLM, ERP, your inbox, WhatsApp, even the PLCs and machines on the floor. No big migration first.

04

Measured in hours saved

If it doesn't save your team measurable time, it isn't finished.

From the founder
"To me, a slow corporate process is just a badly engineered machine. Thuniv exists to fix that machine."
Gokul Kumar · Founder, Thuniv Industries

Gokul trained as a mechanical engineer and has spent years building AI and ML systems at Maruti Suzuki and Renault Nissan. His work there includes LLM document automation, ML design validation and sentiment platforms used by thousands of people.

View portfolio ↗
Early access

Let's build something great.

We work with a small number of design partners in automotive, industrial and engineering services. If one of these problems sounds familiar, get in touch.

Email us LinkedIn ↗