Founders
Kunal Shaw
check_circle Profile complete
Basics
Name
Kunal Shaw
Email
shawkunal032@gmail.com
Phone Number
+919432810879
Age
22
Gender
male
City where you currently live
Bengaluru, Karnataka, IN
Role
What is your title, or if you haven't set it yet, main responsibility?
CEO
What percent equity do you have?
50
Do you have at least 10% equity in the company you're applying with?
yes
Are you a technical founder?
yes
Are you currently in school?
no
If accepted, will you commit to working exclusively on this project for the next year?
yes
Background
Your LinkedIn URL
Education
IIT Roorkee
Nov 2021 - Jun 2025
Btech, Engineering Physics
Shree Jain Vidyalaya
Jan 2019 - Jan 2021
Higher secondary, Physics, Chemistry And Mathematics
Work Experience
ProtonCraft - Founder / CEO
Jun 2024 - Present
Led ProtonCraft from lab concept to TRL-6, pilot-ready prototype producing green hydrogen from refinery H₂S waste at 75% lower energy than conventional routes. Our electrocatalyst outperforms the global benchmark by 8X (validated experimentally). Filed 3 IPs: electrocatalyst composition, system design, and an origami-based electrode architecture. Closed LOIs with India's biggest oil companies, BPCL and IOCL. Raised pre-seed funding, built a 5-person team, and run our own lab.
Hult Prize Foundation - Global accelerator participant (Top6)
Jun 2024 - Aug 2024
At London, ProtonCraft represented India as one of the Top 16 startups globally out of 20,000+ startups at the Hult Prize and emerged as Top 6 global finalists.
IPR Cell IIT Roorkee - Project Lead
Sep 2023 - Feb 2024
Led a team of designers and engineers to ship an internal IP management platform for one of India's top technical institutions, end-to-end digitalization of patent, copyright, and trademark workflows with built-in confidentiality.
Density Exchange - Business Analyst
May 2023 - Jul 2023
Sole technical member embedded in a marketing team. Responsible for integration, tracking, and deployment of analytics and engagement tools (PostHog, MoEngage, GA4).
Energy Storage Laboratory(ESL), IIT Roorkee - Research Associate
Feb 2023 - Mar 2025
I built a deep learning-based BMS for sodium-ion batteries using proprietary experimental data from the lab.
I was youngest researcher on the team working toward India's first MW-scale sodium-ion plant. This is where I first saw the gap between available simulation tools and what real electrochemistry R&D actually needs, the direct origin of BlueDotLabs.
Exquisite - BackEnd Developer(Now this role is dead)
Oct 2022 - Dec 2022
Developed doubt solving platform for our client Step App by Pace IIT & Medical. Coordinated with design and product team to make product using MERN Stack
Social Media
Github URL
Accomplishments
Please tell us about a time you most successfully hacked some (non-computer) system to your advantage.
I legally hacked Indian Banks. I took education loan to cover my all education expenses and then received a merit scholarship as reimbursement, Invested in curated stocks, covered cost of my initial startup and recovered with 100% profit.
Please tell us in one or two sentences about the most impressive thing other than this startup that you have built or achieved.
Built an electrocatalyst that achieved 8X better performance than published literature (200mA/cm² @ 534 hrs vs. industry standard of 25mA/cm²), validated through independent testing. Every professor said it would take at least two years.
Tell us about things you've built before. For example apps you've built, websites, open source contributions. Include URLs if possible.
Multiscale simulation for catalyst synthesis — starts from DFT energies, integrates into kinetics, and predicts what composition you'll get from electrodeposition. Can sweep parameters to find the exact concentration and conditions for a desired result.
Camera-based data acquisition for physical experiments — updates us 24/7 on Telegram, maintains logs, flags if something goes wrong. We use it daily at ProtonCraft.
First startup AlgoEase (Pre AI era) — built a full pipeline to execute trading strategies on live Binance APIs, running simultaneously on 36 coins using parallel threading.
List any competitions/awards you have won, or papers you've published.
-Hult Prize Top 6 global finalist(Top 0.03%)- Selected from 20,000+ startups across 130+ countries.
-Singhal tech Society award- Best overall technical project among 2,614 IIT graduates.
-Winner of Together 5.0 - Indo-Canadian tech startup Competition, selected from 200,000+ applicants.
List any relevant or impressive test scores.
-JEE- 99% percentile out of 1.2M+ candidate(Top-1%)
-All India Rank 541 in WBJEE out of 150K+ candidate
-International Mathematics Olympiad(IMO) regional Rank 52
List any entrepreneurship programs, clubs, or hacker houses you have participated in or are currently participating in.
-Hult Prize Global Accelerator, London, UK (2024) — Completed a 2-month program focused on Pitching, global finance and international business expansion.
-At ESummit'24 Led 120+ team in Entrepreneurship Cell, IIT Roorkee to conclude successful event receiving Footfall of 15k, INR 3M funds and hosted nationally recognized entrepreneurs.
Kamal Utla
check_circle Profile complete
Basics
Name
Kamal Utla
Email
kamalutla@gmail.com
Phone Number
+918179298095
Age
23
Gender
Male
City where you currently live
Bengaluru, Karnataka, IN
Role
What is your title, or if you haven't set it yet, main responsibility?
CTO
What percent equity do you have?
50
Do you have at least 10% equity in the company you're applying with?
yes
Are you a technical founder?
yes
Are you currently in school?
no
If accepted, will you commit to working exclusively on this project for the next year?
yes
Background
Your LinkedIn URL
Education
IIT Roorkee
Nov 2020 - May 2024
Btech, Engineering Physics
Sri Venkateswara Junior College
Apr 2018 - May 2020
High school diploma, Physics, Chemistry And Mathematics
Work Experience
BorderPlus - Founding AI Engineer
Nov 2025 - Present
First AI hire. Built the company's entire AI stack from scratch, owned everything from prompt infrastructure to RAG systems, knowledge base, and agentic workflows.
Designed and shipped the knowledge platform that turns dense regulatory documents into structured, searchable rules — now powering the company's core product.
Built multi-step AI agents that process clinical transcripts, extract tasks, and generate structured notes for downstream healthcare systems.
Architected a multilingual RAG system for grounded Q&A over healthcare literature with inline citations.
Cut production token usage by 15% through prompt orchestration and guardrails.
Convai - Data Scientist - Applied Research Team
Jun 2024 - Oct 2025
Owned the company's entire LLM and NLP stack and supported the team to scale to $3M ARR, after my manager left - became the sole person responsible for everything from model serving to fine-tuning to evals.
Deployed and served large models (OpenAI, Anthropic, Gemini, LLaMA) on custom cloud infra with vLLM, including guardrails and tool calling for production reliability.
Fine-tuned large multimodal models (LLaMA-3-70B, LLaMA-3.2-90B-vision) on custom data pipelines to improve output quality.
Built a novel hybrid semantic + BM25 chunking method that cut RAG latency by 27% and reduced memory usage 3–4x.
Prototyped multi-agent systems (director-orchestrated group chats), browser agents for task automation, and real-time chat plugins for contextual QA over live websites.
Convai - Machine Learning Intern
Feb 2024 - May 2024
Built a prompt-engineering framework that modeled Big 5 personality traits in LLMs to make AI characters feel consistently human.
Fine-tuned small open-source models (Gemma-2B, Phi-3-3.8B) on synthetic data pipelines I designed for emotion perception and contextual response — plus evaluation frameworks to measure character consistency.
Energy Storage Lab, IIT Roorkee - Undergraduate Research Assistant
Aug 2023 - Apr 2024
Worked on building intelligent Deep learning models for predicting state of charge of sodium ion batteries. Built a custom dataset for Indian battery conditions in electric vehicle
DeGroote School of Business - McMaster University - MITACS Globalink Research Intern
May 2023 - Aug 2023
Built non-linear optimization models in Python to determine optimal EV battery recycling center locations across the Greater Toronto Area, balancing financial and environmental costs.
Banao Technologies - Technical Product Manager Intern
Aug 2022 - Oct 2022
Managed a team of developers, testers and designers to successfully deliver the product to our clients. Handled client meetings and closely coordinate with all the stakeholders to effectively plan the sprints, establish timelines and tasks.
IIM Bangalore - Summer Research Intern
May 2022 - Jul 2022
Conducted empirical research on the correlation between founder human capital metrics and startup funding success rates.
Developed and implemented a multi-factor analysis framework examining educational backgrounds, professional experience, and funding outcomes.
Performed cross-cultural analysis of startup ecosystems, evaluating the impact of governance structures and regulatory frameworks on funding dynamics.
Social Media
Personal website
Github URL
Accomplishments
Please tell us about a time you most successfully hacked some (non-computer) system to your advantage.
In a national quantum computing hackathon, most teams treated it like a PhD defense—dense algorithms, heavy theory, and demos that barely worked.
I suspected the judging system didn't actually reward "most complex," it rewarded "most understandable + working." So I optimized for that instead.
I put together a team with a quantum dev, backend + frontend engineers, and—slightly irrationally - someone with a music background. We built a simple "quantum music generator": not the deepest project, but one you could see, hear, and understand in 30 seconds.
While other teams were explaining what their systems would do, we showed something that already did.
We won.
That experience stuck with me - I don't take systems at face value anymore. I try to figure out what they actually reward, and then play that game deliberately.
Please tell us in one or two sentences about the most impressive thing other than this startup that you have built or achieved.
Faced with a failing RAG system at Unilever, I redesigned chunking and hybrid retrieval-reducing memory 3–4x and latency by 27% - and validated it on a 350M+ vector pipeline I built alone.
Tell us about things you've built before. For example apps you've built, websites, open source contributions. Include URLs if possible.
A lot of my work has revolved around building agentic frameworks using LLMs.
I worked on a previous Idea where I'd built an automated LLM evaluation suite, which could generate testcases and evaluate llm pipelines with ease. I have made some fun projects where I built a speaking coach for people to approach their preferred gender partners confidently. I created a smart travel planner which would give an itinerary for a trip based on user's preferences and target locations by scraping forums and the web in general.
In my current company I have built apps where we can generate L&D roleplay scenarios for nurses with actionable outcomes, based on some nursing books and legal standadrd. Setting up a complex agentic RAG pipeline along with a heavy knowledge base for german nursing usecases.
At my previous company I have worked on building a full stack AI powered group chat app where multiple AI personas talk with us, I have built webRTC based browser agents which can understand webpages and perform tasks in it. Also worked on finetuning LLMs for complex emotional and physical action capabilities, and towards the abliteration of LLMs.
List any competitions/awards you have won, or papers you've published.
1. IIT R Encore award 2023 for all round excellence among 1200 students in my batch.
2. Mitacs Globalink Research Fellowship: selected for a research fellowship in Canada among 1000s of global applicants.
3. Winner, IBM |Qiskit Fall Fest 2021: National Level Quantum Computing hackathon hosted in IIT Roorkee
4. Runner up, Startbucks - national level startup pitching competition hosted by IIT BHU
5. Winner, Enact-a-thon - Social Startup pitching competition, hosted by IIT Roorkee
6. Defining Innovation Award, Social Budgeting Convention 2023 - hosted by TIET, a social business case study competition
List any relevant or impressive test scores.
- among the top 10% in JEE advanced - One of the toughest competitive exams
- among the top 5 in my department at IIT Roorkee
List any entrepreneurship programs, clubs, or hacker houses you have participated in or are currently participating in.
- Worked in the Social Entrepreneurship club in my college, worked and led projects involving rural tourism, waste recycling and sustainable product devlopment.
- Participated in Tides Winter cohort 2020 (Incubation cell of IIT Roorkee), worked with my team on ideating products around smart masks and protection kits for covid.
How long have the founders known one another and how did you meet?
At Engineering Physics at IIT Roorkee, we've known each other for five years and spent our final year working together on building an intelligent Battery Management System (BMS) in the ESL lab which worked for sodium ion cells.
Who writes code, or does other technical work on your product?
Kamal writes the code and Kunal make scientific modules which agent uses to orchestrate simulations.
Are you looking for a cofounder?
No
Founder Video
Company
Company name
BlueDot Labs
Describe what your company does in 50 characters or less.
Physics simulations for energy R&D, 100x faster
Company URL
Demo Video
Product link
Login credentials for the product
demo@bdlabs.ai / pimono@032
What is your company going to make?
We've built an AI-native platform that reduces complex battery simulations from 7-8 days to under 30 minutes witnessed by e-TRNL energy, a battery company.
The AI agent plans, builds, and runs scientific simulations from plain text. It converts a text prompt into an editable flowchart and clarifies constraints with the researcher before running the physical model on our compute.
Researchers can instantly run parameter sweeps and visualize results, condensing weeks of complex scientific modeling into minutes of prompting so researchers can focus purely on the science.
Where do you live now, and where would the company be based?
Bengaluru, India / San Francisco, USA
Explain your decision regarding location.
SF gives our AI product direct access to the network, high-paying customers, top talent, and deep-tech investors.
Progress
How far along are you?
In 3 months, we shipped V0, obtained paid contracts with two B2B battery companies (e-TRNL Energy, Arkle Energy) which executes with V1 launch, and onboarded three Tier-1 institutions (Caltech, TU Munich, TU Dresden) as design partners. We're now iterating to V1 based on our customers' feedback to deliver on those contracts.
How long have each of you been working on this?
3 months. Kunal full time, Kamal joins full-time June 4th.
Kunal ran ProtonCraft for 1.5 years. BlueDotLabs is the internal tool we built out of desperation. Realized every deep-tech team has the same problem. ProtonCraft is a local maxima. We're going for the bigger one.
What tech stack are you using?
Backend: Python. Coding agent server: TypeScript (PiMono — same engine behind OpenClaw). DB: PostgreSQL. Frontend: Next.js. Flow builder: React Flow.
Build with Claude Code + OpenCode/GLM, customized with Playwright skills for E2E testing. UI/UX starts in Google Stitch.
Coding agent session
Attach a transcript you're proud of
Parametric 3D web editor for e-TRNL unit cell
Session: ses_23690f7deffeNa0pUm1Hql7pB4 | Duration: 55.6 hrs | Messages: 951
Patent-faithful 3D mesh of the e-TRNL 3DEA battery unit cell — cathode pins embedded in anode matrix, inspired by origami/pin-array patent.
hex_R=1.95mm,H=4.0mm,d_pin=0.30mm,t_sep=0.025mm- 61 pins in 4-ring HCP lattice
Built brainstorming → design → full Three.js in single session. Full transcript on request.
Are people using your product?
Yes
How many active users or customers?
Paying: 2 (e-TRNL Energy, Arkle Energy — $500/seat/month, activating with V1).
LOIs: Indi Energy, Onnes Cryogenics.
Design partners: 3 researchers at Tier 1 universities.
Pipeline: BPCL (~$50B revenue) on waitlist. Vetri Labs demo scheduled.
Revenue?
No
Why did you pivot from ProtonCraft?
BlueDot Labs is the direct solution of the problem we faced daily at ProtonCraft for 1.5 years. We desperately needed AI simulation tools to pace up R&D, so we started building it.
At ProtonCraft, we built 8x better technology and reached TRL-6. But velocity was bottlenecked by slow customers (oil refineries) and heavy regulations.
Core learning: relying on legacy industries to validate your tech kills momentum. We pivoted to software so we can own the entire product cycle and move fast.
Any incubator/accelerator participation?
No
Idea
Why did you pick this idea? Domain expertise?
We want to build a faster research ecosystem where the focus is science, not labor. We picked this because we've lived it — watching researchers lose weeks wrestling with clunky simulation tools. With current AI models, it's possible to do now what wasn't possible even 6 months ago.
Both have domain expertise. Kunal lived the exact problem at ProtonCraft. Kamal scaled two startups as founding AI engineer with simulation interest.
Customers like e-TRNL and Indi need not just software licenses but a specialist whose only work is translating science into software GUI.
Competitors? What do you understand that they don't?
COMSOL, ASPEN, ANSYS, SimScale — black boxes with decades-long update cycles. Godela, PhysicsX — mostly mechanical engineering, not energy R&D.
Researchers can't see the physics, can't swap solvers, can't afford the licence. 80% of better open-source solvers sit unused because integration took weeks — until now.
AI is finally good enough to orchestrate domain-specific simulations correctly. We're building at that inflection point — modular, transparent, pay-per-use. Starting with energy R&D where wrong answers cost years.
How do/will you make money?
6 B2B customers in Bangalore spending $150K–$300K/year on licenses + simulation engineers. We replace at $20K–$25K/year, $147K ARR, 6 weeks to close post-launch.
US launch 6 months: 2–3× deal sizes. Month 9: $647K ARR, 13 customers.
3,000 teams globally = $927M ARR opportunity. Just 5% of $20B simulation software market, growing 10.4% annually to $36B by 2030.
Category?
B2B SaaS
Other ideas you considered?
Scientists should only do science.
Physical AI to automate all manual labor in wet labs — starting with electrochemistry.
AI-powered research lab that runs autonomously — a second Bell Labs, but limitless.
Equity
Have you formed any legal entity?
No
Planned equity breakdown
Equal split.
Kunal — CEO, 50%
Kamal — CTO, 50%
Taken any investment?
No
Currently fundraising?
No