Spearhead practice
Fractional CTO for AI Startups
I run ColadAI, a multi-LLM platform, and my Masters research was in cybersecurity, which means I have spent the last two years living inside the exact problems that break AI startups. Not the demo. The part after the demo.
Most AI companies I talk to have a working prototype and no idea what it costs to serve, no way to tell whether a prompt change made things better or worse, and a security surface nobody has looked at. That is the job I am hired to do.
What I'm hired to own
The decisions that are cheap now and expensive later
Your inference costs will decide whether you have a business
A feature that costs $0.40 per user session is fine at 200 users and fatal at 50,000. I model unit economics per feature before you ship it, then bring the number down: routing cheap requests to small models, caching aggressively, batching, and cutting context that is not earning its place. This is usually the single highest-leverage thing I do in month one.
You have no evals, so you are shipping on vibes
If you cannot answer "did that prompt change make the product better?" with a number, you are guessing. I build the eval harness: golden datasets, regression suites in CI, and offline scoring before anything reaches production. Prompt and model changes stop being coin flips.
Prompt injection is a real vulnerability, not a thought experiment
Any LLM feature that reads untrusted input and can call a tool is an injection target. I have written publicly on this. I define the trust boundaries, sandbox tool calls, scope credentials down, and put the controls in before your first enterprise security questionnaire arrives, not after it fails one.
Single-vendor dependency is a real risk you can price
Rate limits, deprecations, and pricing changes are not hypothetical. I put an abstraction between your product and any one provider, with tested fallbacks and a measured quality delta per model, so a vendor incident is a degraded hour instead of a dead product.
AI-written code accumulates debt faster than it saves time
Teams shipping mostly-generated code hit a wall around month six: nobody understands the codebase, tests are decorative, and every change breaks something adjacent. I set the review standards and architectural boundaries that keep velocity from collapsing.
How an engagement runs
Week one, month one, month three
Week 1
I find out what is actually true
Architecture review, codebase read, and honest conversations with every engineer you have. You get a written assessment at the end of it: what is solid, what is risky, what is going to break, and what it costs to fix each one. No slide deck, just a document you can act on.
Month 1
The bleeding stops and the roadmap becomes real
I fix or contain the things from week one that were actually urgent, and rebuild the technical roadmap so it has dates you can believe. If you are hiring, the role is written and the interview loop exists. If you are pre-launch, scope gets cut, usually by a lot.
Month 3
It runs without me in the room
Ship cadence is predictable, the team knows the standard, security and infrastructure are at the level your customers expect, and the decisions are documented. If the engagement is a bridge to a full-time CTO, this is where I start hiring my replacement.
Why me, for this specifically
Experience, not a capabilities deck
Plenty of people can advise on ai startups. Fewer have shipped in it, been wrong in it, and had to fix it with their own name on the commit. Here's what I'm bringing.
- Founder of ColadAI, a production multi-LLM platform serving real traffic
- M.Tech in Cybersecurity, NIT Kurukshetra, with 30+ academic citations
- Published on prompt injection and AI-driven threat detection
- Six ventures built 0 to 1 across three continents
Pricing
Priced at $100 an hour. Not counted at it.
I value my time at $100 an hour and the retainers are sized from that, but the hours are a floor I commit to rather than a cap I bill against. If a week needs more, it gets more. A full-time CTO in the US runs about $260,000 a year before equity; I'll do that job for $60,000 at 150% commitment. Three-month minimum, then month to month.
It does not all have to be cash. The split is negotiable, from a small equity component up to roughly half and half, depending on your stage and how long we expect to work together.
Technical Advisor
$1,500
per month
Minimum 15+ hours a month
A senior technical brain on call. For founders who mostly need the big decisions checked rather than someone in the codebase.
Fractional CTO
$5,000
per month
Minimum 50+ hours a month
Around 50 hours a month as your acting CTO, usually two days a week. I own technical direction, and the roadmap ships because someone senior is accountable for it.
Embedded CTO
$8,000
per month
Minimum 80+ hours a month
Around 80 hours a month, usually three days a week, hands on the keyboard. Everything above, plus I build the hard parts myself instead of delegating them.
Questions
AI startups, specifically
Do you work with pre-product AI startups?
Yes, and it is often the cheapest time to hire me. Deciding what not to build (which model, whether you need fine-tuning at all, whether the feature is even economically viable at scale) is worth more than fixing it in month nine.
Can you help with technical due diligence for an AI round?
Yes. Investors increasingly ask hard questions about model dependency, data rights, eval rigour, and security posture. I prepare the technical narrative and fix what will not survive scrutiny before the diligence call, not during it.
We are using an AI coding agent for most of our code. Is that a problem?
Not inherently. I use them too. It becomes a problem when there is no senior review layer and no architectural boundaries. That is a solvable process problem and one of the first things I set up.
More general questions about cost, commitment, and how this compares to a full-time CTO are answered on the main fractional CTO page.
Earned, Not Claimed
What the founders and teams I build with say
I had the pleasure of working with Kunal while building out my MVP and I have to say that it was pleasure to both interact and work with him and his team. As a first time founder, I had a lot of self doubt and internal ruptures, however Kunal's calm, never-say-die and never-panic attitude often helped bring me back to center. I can't vouch for him enough.
Anand Tahiliani
Founder, Debaser Technologies
I had the pleasure of working closely with Kunal during my time as a Product Manager at DhunGuru. As our CTO, Kunal and his team at Panicle Tech worked closely with me across the entire product journey, from building and launching our mobile apps to developing and maintaining our web platforms. His technical expertise, ownership, and willingness to collaborate made a real difference, and I genuinely enjoyed working with him throughout the journey.
Anirudh Joshi
Product Manager, DhunGuru
Kunal helped us architect Mosler's core platform, connecting smart lock hardware, our cloud dashboard, and PMS integrations into one reliable system. The foundation he helped design in those early days still scales with us today.
Pranav Kapoor
Co-Founder, Mosler
Building in ai startups? Let's talk.
Thirty minutes. Tell me what you're building and where it's stuck, and I'll give you a straight answer on whether I can help.
Book a 30-min callSend me a message