// PODCAST

Deeptech Investing in India: How a VC Evaluates Startups | Manoj Agarwal, Seafund Ventures

Manoj Agarwal of Seafund Ventures explains how an early-stage deeptech VC sources, filters and diligences deals, and what founders get wrong about compliance and runway.
Manoj Agarwal, Seafund Ventures on The Builders Club Startup Founders Podcast
Advertisement

A deeptech VC can see 300 to 500 deals a month and invest in only a handful. What separates the few that make it through?

Manoj Agarwal, Managing Partner at Seafund Ventures, walks through deeptech investing from the inside: how deals are sourced and filtered, how diligence works and the mistakes founders keep repeating.

What is deeptech investing, and how does Seafund Ventures approach it?

Seafund Ventures is an early-stage fund investing at seed and pre-Series A in deeptech and enterprise B2B technology. Manoj says deep tech grew out of an earlier focus on technology: the capital needs are lower than in consumer businesses, and the government now backs areas such as semiconductors, space and EV.

His approach is to expect every company to be a potential fund-returner, filter hard, run real diligence and support founders after the cheque is written.

  • Heavy filtering. Roughly 300 to 500 deals arrive each month, narrowed to 30 or 40 for conversations.
  • Gut gets you to the table, numbers close the deal. Every investment goes through diligence.
  • Compliance early. Messy company formation and tax filings can sink funding or an acquisition.
  • Careful with hype. Seafund backs AI as a tool inside real businesses rather than AI for its own sake.

What is Manoj Agarwal’s background and why did he start Seafund?

Manoj studied computer science at BMS College of Engineering in Bangalore and graduated in 1993. He built an IT infrastructure business and a multimedia training company, ran a chain of cyber cafes during the dot-com years and later an enterprise ISP. He also had stints in food processing and floriculture.

After a part-time management programme, he began advising companies and making angel investments. That led to a small angel fund that began investing in 2018 and backed about 14 companies, then Fund 2, which had its first close at the end of 2022.

Why does Seafund invest in deeptech?

Manoj says technology is often only an enabler in consumer businesses that need large capital. Enterprise and deeptech startups can survive hard phases with less capital.

Government focus has also widened the opportunity. He points to support for semiconductors and space, and says non-software engineering talent now has more routes into startups.

  • Robotics, such as a portfolio company building robots to eliminate manual scavenging and for rehabilitation.
  • EV infrastructure, such as a robotic battery swapping station for cars.
  • Drone logistics, focused first on healthcare.
  • AI SaaS applied to areas like fraud detection in insurance.

How does a VC filter hundreds of deals a month?

Seafund uses initial filters to cut 300 to 500 monthly deals to around 30 or 40 conversations. Referrals help because they come with some validation, and Manoj says they can come from other VCs, chartered accountants and even family.

He rates gut as a starting point only. A founder he met two years earlier in Abu Dhabi, for example, became an investment once he began raising and the firm completed its evaluation.

  1. Screen against stage, sector and valuation fit.
  2. Check whether a follow-on investor is likely to write the next, larger cheque.
  3. Test whether revenue can reach meaningful scale.
  4. Run a deep dive on compliance, finances, product and technology.

What does due diligence uncover in early-stage startups?

Manoj says most issues are fixable compliance gaps, and the firm helps founders correct them. Deals have been dropped when the technology claims did not hold up or the business plan, once corrected, showed the round could not even fund year one.

He stresses this is rarely deliberate lying. Often founders simply do not understand their numbers or overestimate how different their product is.

What mistakes do founders make on compliance and runway?

His favourite theme is compliance and financial management from day one. He has seen founders spend crores building a product with no company to account for it, and an acquisition that failed at diligence over such gaps.

He also warns that terms like SAFE notes do not translate neatly to Indian law, and that ESOPs are often created without understanding them.

  • Form the company properly and record share allotments and ROC filings.
  • Pay TDS, PF and GST on time, because small arrears can haunt a deal later.
  • Know your burn and bank balance well before runway drops below three months.
  • If revenue is steady, consider cutting costs to break even rather than shutting down.

How should founders think about AI hype and FOMO?

Manoj sees AI as a technology here to stay, but warns of noise. Seafund prefers companies using AI as a tool within a larger product. He gives enterprise examples where AI handles most of an accounting or compliance task and a human reviews the rest.

He describes sectoral FOMO, such as rocket companies that raised heavily, and cautions that high valuations can leave founders diluted.

How much does Seafund invest, and what is its plan?

Fund 2 plans to back up to about 20 startups, with a small slice reserved for follow-ons. Around seven deals are done, with 11 to 13 more planned over roughly 18 months.

First cheques go up to around half a million dollars, sometimes 250,000 to 300,000 dollars, in areas such as semiconductors, aerospace, mobility, materials and AI SaaS.

Further reading and sources

Frequently asked questions

What do deeptech VCs look for in early-stage startups?

Manoj Agarwal looks for real innovation, a path to meaningful revenue and a founder who can raise the next round. Compliance and financial hygiene also matter, since gaps can derail investment.

How big are Seafund Ventures’ first cheques?

Seafund writes first cheques up to about half a million dollars, and sometimes 250,000 to 300,000 dollars. The size depends on stage and valuation, and it covers pre-seed to seed.

Should an individual invest in startups directly as an angel?

Manoj says angel investing is close to a full-time job. If you cannot dive deep, he suggests investing through a fund or alongside someone who does it full time.

Conclusion

Manoj’s message is practical. Raise from investors who understand your space, keep your compliance clean from the start and know your burn long before it becomes urgent.

Before approaching a deeptech fund, check that your cap table, filings and numbers would survive a deep diligence review.

Want more conversations like this? Join The Builders Club, our community of 35,000+ founders and CXOs, and explore more episodes of the Startup Founders Podcast.

Never miss an episode

Join 35,000+ founders and CXOs in The Builders Club community.

Join the Community
Advertisement
Advertisement
Facebook
X
LinkedIn

Why the Self-Serve PLG Playbook Fails with Indian Enterprise CXOs (And the High-Touch Assisted Model That Actually Wins)

Self-serve PLG works when the user is also the buyer. In Indian enterprises, they almost never are. A product-led funnel can get your tool into a team’s hands, but it rarely gets a CXO to sign. The model that wins is a high-touch, assisted motion that uses product usage as proof, not as the whole sales strategy. Here is why the standard playbook stalls in India’s large organisations, and what to run instead. Why does the self-serve PLG playbook fail with Indian enterprise CXOs? Product-led growth, as described by pioneers such as OpenView, assumes that users discover, try and adopt the product themselves, and that revenue follows usage. That assumption holds for tools with low risk, low price and individual buyers. Enterprise buying in India breaks all three conditions at once. The user is not the buyer. An analyst may love your tool. The budget sits with a CIO, CFO or business head who has never opened it. Risk is organisational. A CXO signs against security, vendor-continuity and reputational risk, none of which a free trial answers. Trust is personal. Decisions often move on credibility, references and conversations with someone they respect, not on a sign-up flow. Does a free trial or freemium plan ever work in Indian enterprises? It can work for teams, departments and SMB-sized units, where one manager has a card or small discretionary budget. It rarely scales to the top accounts without a human who turns usage into a business case. What does the high-touch assisted model actually look like? Assisted does not mean old-fashioned. It means every stage of the product experience is paired with a person who removes a specific obstacle. Stage Pure self-serve PLG Assisted model for Indian enterprises Discovery Organic search, product trial Warm introduction, peer room, referral Evaluation Self-guided trial Guided pilot with named success criteria Decision User upgrades with a card Business case built for the CXO and the committee Contracting Click-through terms Negotiated MSA, security review, procurement Expansion In-app prompts Account lead plans rollout across divisions How do you turn product usage into a signed enterprise contract? Use the product as the evidence, and a person as the translator. A practical sequence: Find your champion. Identify the user who gets the most value and ask them to explain the problem in their own words. Quantify the result. Convert usage into time saved, errors avoided or revenue influenced, in a one-page summary the CXO can read in two minutes. Map the buying committee. Ask your champion who else must say yes: IT, security, finance, legal, and the business owner. Run an executive conversation. A founder or senior lead speaks to the CXO about outcomes and risk, not features. Offer a scoped commitment. A defined pilot or first-division rollout with clear exit and expansion terms reduces perceived risk. When should you still invest in self-serve? Do not abandon it. Use self-serve as the top of the funnel and as a signal engine. Track which accounts have several active users, which teams invite colleagues, and which admins explore security settings. Those are the accounts your sales team should call first. Think of it as product-qualified leads feeding a human-led close. Many global companies now run exactly this hybrid, and it fits Indian enterprise buying better than a pure self-serve approach. FAQ Is PLG dead for B2B SaaS in India? No. PLG is useful for building awareness, usage data and champions inside accounts. It fails as the only motion for large enterprises, where committees, security reviews and relationships decide the contract. A hybrid model works better. What is a product-qualified lead (PQL)? A PQL is an account or user whose in-product behaviour suggests buying intent, such as multiple active teammates or admin-level exploration. Sales teams use PQLs to prioritise who to approach with a human conversation. How many people does an assisted enterprise sale need? It depends on deal size, but you typically need one senior owner for the relationship, one technical resource for security and integration questions, and a customer success lead after signing. Founders often cover the first two roles in early deals. Conclusion PLG gets you into the room, but it does not get a CXO to sign. In Indian enterprises, winning sales pair product evidence with human trust: a champion, a business case, a guided pilot and a senior conversation. Building an assisted enterprise motion? Meet founders and CXOs who have done it at The Builders Club, and bring your toughest sales question to our next roundtable. Join The Builders Club. About the author: Sohail Khan is Founder and Director of The Builders Club, a professional community of 35,000+ founders and CXOs across India, UAE, Singapore and the US.

Read More »