Seed Innovations Pivots to Robotics and AI, Deploys Capital in Feather and Fieldwork

Seed Innovations final results show an AI and robotics pivot, but the portfolio is still mostly legacy biotech

Seed Innovations has used its latest final results to draw a clear line under its old identity and pitch itself as an AIM vehicle for high-growth robotics and AI ventures. That is the big headline here. The catch is that, at 31 March 2026, the actual portfolio was still overwhelmingly made up of legacy life sciences, biotech and cannabis-related holdings.

So this is really a transition story rather than a finished transformation. For retail investors, that matters because you are not yet buying a pure-play robotics portfolio – you are buying a company trying to rotate into one.

Seed Innovations key numbers: NAV, cash, portfolio value and discount

Metric	31 March 2026	31 March 2025
Investment portfolio value	£8.4 million	£8.3 million
Net asset value (NAV)	£11.0 million	£11.8 million
NAV per share	5.88p	6.10p
Cash and cash equivalents	£2.8 million	£3.4 million
Total comprehensive gain/(loss)	(£609,000)	£367,000
Basic gain/(loss) per share	(0.32p)	0.19p
Share price discount to NAV at year end	56.6%	73.77%

NAV means net asset value – essentially the value of the portfolio and cash, less liabilities. The discount to NAV tells you the market is valuing SEED at far less than its stated assets, although the gap has narrowed materially from last year.

Why the Feather Robotics and Fieldwork Robotics investments matter

The strategic pivot only really became tangible after the year end. SEED invested US$1,000,000, about £740,000, into Feather Robotics in May 2026 and £300,000 into Fieldwork Robotics in June 2026.

That is important because it gives shareholders actual exposure to the new theme rather than just boardroom talk. Without those deals, this announcement would have read more like a rebrand than a strategy.

Feather Robotics gives Seed Innovations exposure to “physical AI”

Feather Robotics is described as a US-based applied AI and robotics company building modular humanoid robotic systems for industrial environments. SEED invested via a SAFE, or Simple Agreement for Future Equity, which is a funding instrument that converts into equity at a later qualifying funding round or liquidity event.

The SAFE has a valuation cap of US$60 million. That can be attractive if Feather grows quickly, but it also means investors do not yet have ordinary equity in hand. It is early-stage and high-risk by definition.

Fieldwork Robotics adds a real-world agricultural automation angle

Fieldwork Robotics is developing an autonomous raspberry harvesting robot and SEED subscribed for A1 preferred shares at an implied post-money valuation of £8.3 million. Trials are underway in Australia, the UK and Portugal, with expansion into the US planned.

This one looks easier for retail investors to grasp. Labour shortages in agriculture are a genuine commercial problem, so if the product works reliably, the addressable market could be meaningful.

Seed’s financial results were weaker, but that is not the main point of this update

The hard numbers were not especially exciting. SEED posted a total comprehensive loss of £609,000 for the year, versus a £367,000 gain in 2025.

Total investment gain fell sharply to £185,000 from £979,000. At the same time, total expenses rose to £895,000 from £744,000, with legal and professional fees up to £146,000 from £53,000 and new investment consultant fees of £69,000.

In plain English, the portfolio did not revalue strongly enough to cover the running costs. That is not unusual in an investment company mid-transition, but it does underline why execution from here matters so much.

Cash has fallen, and that limits how fast Seed Innovations can build the new robotics portfolio

Cash and cash equivalents fell to £2.8 million at the year end from £3.4 million. More importantly, after the post-period investments and ordinary operating expenditure, cash reserves were approximately £1.2 million as at 19 June 2026.

That is enough to keep moving, but it is not endless firepower. My read is that SEED can make selective follow-on bets, but if it wants to build a broader robotics and AI portfolio at pace, it may eventually need either exits from the legacy book, a rerating in the shares, or fresh capital.

Legacy portfolio holdings still dominate the valuation – and that cuts both ways

At 31 March 2026, the largest holdings were Avextra AG at £3.2 million, Juvenescence at £2.4 million and Clean Food Group at £1.7 million. Together, those three positions made up 66.8% of NAV.

That concentration matters. On the positive side, if one or two of these businesses deliver strong commercial or funding milestones, SEED’s NAV could move nicely. On the negative side, shareholders are still exposed to sectors that are not the new robotics and AI focus.

Clean Food Group remains a conviction holding

SEED also put a further £260,000 into Clean Food Group after the year end as part of a £4.5 million financing via a convertible loan note, or CLN. A CLN is debt that can convert into equity later, and this one carries a 12% annual coupon and matures in August 2027.

That follow-on investment says the board still sees value in selected legacy assets. It is sensible not to dump everything old just to make the new story cleaner.

Inveniam is the clear yellow flag in the portfolio

The most cautious wording in the whole results relates to Inveniam. Management disclosed potential near-term funding pressures, is cutting costs, seeking to realise certain assets and undertaking a fundraise.

SEED explicitly said these actions may lead to a review of the carrying value at the interims. That is a polite way of saying there could be a markdown coming.

The narrowing discount to NAV is encouraging, but the market still wants proof

SEED’s shares were trading at a 56.6% discount to NAV at 31 March 2026, improved from 73.77% a year earlier. The company says that has reduced further to about 50% at the current share price.

That is a genuine positive. A discount that wide usually screams distrust – either of valuations, strategy, liquidity, or all three. The fact it has narrowed suggests the market likes the new direction and the arrival of Jim Mellon as Non-Executive Chair, but a 50% discount still shows a lot of scepticism remains.

Governance and related-party fees are worth watching closely

One detail investors should not ignore is Hoid.ai, the investing consultant appointed in December 2025. Hoid.ai is controlled by an entity controlled by Jim Mellon, and both Jim Mellon and Denham Eke are directors of Hoid.ai.

SEED charged £69,390 for Hoid.ai in the year, all payable at the reporting date. Hoid.ai is also entitled to a performance-related fee of 15% of the increase in audited NAV per share above a 4% hurdle rate. None of that is automatically bad, but with related-party arrangements, shareholders should expect proper scrutiny and clear evidence of value added.

My view: this is a more interesting company now, but it is still an execution story

I think these results are more positive than negative overall, but not because of the historic financials. The real improvement is strategic. SEED now has a clearer identity, two relevant post-year-end robotics investments, and a board that appears better aligned to the new plan.

The negatives are obvious too. The company is still carrying legacy assets, cash has come down to around £1.2 million post period end, costs are not trivial, and there is at least one portfolio name – Inveniam – that may face valuation pressure.

For me, the investment case boils down to this: if SEED can turn its AI and robotics theme into a disciplined pipeline of investable, early commercialisation businesses, the current discount could look too harsh. If it cannot, this risks remaining a small investment company with an attractive story but limited follow-through.

That makes the next 6 to 12 months crucial. Investors do not need more grand theory about robotics. They need proof in the portfolio.