The AI Layer That Could Make London's Taxis Profitable

The Croydon-based company is using its own electric ride-hailing fleet as a real-world testbed for software it wants to sell to autonomous vehicle operators.

About Odysse Ltd.

Published

For a company that says its AI is the product, Odysse Ltd. owns a lot of cars. About 30 electric vehicles, as of last October, The fleet is not an asset play, but a live data engine. It’s a pragmatic, capital-intensive way to build a software business for a market that doesn’t quite exist yet: the orchestration of autonomous vehicle fleets at city scale.

Founded in 2020, the Croydon-based company is running a dual-track operation. On one track, it operates a premium, all-electric ride-hailing service, completing over 150,000 trips and 1.3 million zero-emission miles by April 2026 [Zenzic, April 2026]. On the other, it is developing the underlying intelligence platform, using reinforcement learning to dynamically position vehicles and process some 600,000 daily demand signals [Odysse website, retrieved 2026]. The bet is that the operational rigor and proprietary models honed on today’s human-driven fleet will become the essential operating system for tomorrow’s autonomous ones.

From Compliance Officer to Fleet Orchestrator

The founder’s path is unconventional for a mobility tech startup. Anant Prakash, Odysse’s CEO, spent years as a secretary, general counsel, and compliance officer for public companies like Synthomer PLC [Bloomberg, retrieved 2026]. That background in governance and regulated operations is less about coding and more about building a company that can navigate complex frameworks,a relevant skill for a business that must comply with Transport for London (TfL) licensing, vehicle maintenance, and employment rules for its drivers.

The technical heavy lifting has been led by Aditya Nigam, who joined as Head of AI, Data & Technology about a year ago. In a recent post, Nigam outlined a methodical build, emphasizing architecture over flashy models. His team focused first on scalable data pipelines and cloud-native infrastructure, then on proprietary machine learning models for driver positioning and in-house reinforcement learning systems [LinkedIn, Oct 2026]. The stated goal is to move from predicting demand to making operational decisions, a shift that reportedly drove a 60%+ increase in revenue per hour versus the ride-hailing industry average [Odysse website, retrieved 2026].

Traction: Contracts and Capital

The company’s progress is measured in vehicles, miles, and, critically, commercial agreements. Odysse reported approximately £1 million in annual revenue as of October 2025 and claims to have generated a similar figure in 2024 [UK Investor Magazine / Acast, Oct 2025] [Moving People, retrieved 2026]. Its most significant commercial milestone is a five-year contract with European mobility platform Bolt, providing a stable demand channel for its fleet services [Moving People, retrieved 2026].

Funding has followed, though the cap table remains opaque. The company raised a $2.5 million seed round in July 2025 and had raised at least £1.6 million by the following September [CB Insights, retrieved 2026] [LinkedIn, Sep 2025]. More strategically, it secured UK government-backed funding in early 2026 specifically to develop an operating blueprint for autonomous private-hire services in London [Zenzic, April 2026]. This points to a validation of its long-term thesis by a stakeholder less concerned with ride-hailing margins and more with future urban mobility infrastructure.

Metric Figure Date / Period Source
Fleet Size ~30 vehicles October 2025 UK Investor Magazine / Acast
Total Journeys 210,000+ Website snapshot (2026) Odysse website
Zero-Emission Miles 1.9 million+ Website snapshot (2026) Odysse website
Annual Revenue ~£1 million 2024 Moving People
Key Partnership 5-year contract Not dated Moving People

The Long Road to Autonomy

The company’s ambition is starkly quantified: 3,000 vehicles and £100 million in revenue by 2030 [UK Investor Magazine / Acast, Oct 2025]. The path there is fraught with execution risk that has little to do with AI models. Scaling a capital-intensive fleet operation while simultaneously developing a software platform requires continuous financing and operational perfection. The current model relies on employed drivers using company-provided EVs, a structure with significant payroll and asset overhead [Odysse website, retrieved 2026].

The competitive set is also evolving. Today, Odysse competes with other premium ride-hailing services and fleet management tools. Its near-term advantage is the vertical integration of its own fleet as a controlled proving ground. Tomorrow, the realistic competitors are the emerging stack providers for autonomous vehicle operations,companies like Motional, Waymo, and a host of startups building AV fleet management software. Odysse’s answer is that its models are being trained on the messy reality of London’s streets today, not in simulation.

The Real Customer Profile

For now, the ideal customer profile is clear: it’s Odysse itself. The company is its own first and most demanding software client, a common pattern in vertical SaaS where the product is born from direct operational pain. The next logical customer is another fleet operator, human or autonomous, that needs to maximize utilization in a dense, regulated urban environment like London. The five-year Bolt deal suggests the model has appeal for platforms seeking reliable, premium-capacity partners.

The realistic competitive landscape for its core AI product isn’t other ride-hail apps, but the internal engineering teams at large AV developers and the enterprise software vendors aiming to serve them. Odysse’s wedge is its owned operational data and its government-backed mandate to draft London’s AV blueprint. The next twelve months will be about proving that the intelligence built for 30 cars can convincingly scale, both within its own growing fleet and, eventually, on the dashboard of a partner who doesn’t want to build this stack from scratch.

Sources

  1. [UK Investor Magazine / Acast, October 2025] Targeting 100x Growth Through Ride-Hailing AI Optimisation | https://shows.acast.com/uk-investor-magazine-flash-briefing/episodes/targeting-100x-growth-through-ride-hailing-ai-optimisation-w
  2. [Zenzic, April 2026] Odysse secures funding to develop blueprint for autonomous private hire services in London | https://zenzic.io/news/odysse-secures-funding-to-develop-blueprint-for-autonomous-private-hire-services-in-london/
  3. [Odysse website, retrieved 2026] Odysse - Fleet Orchestration Intelligence | https://www.ody.tech/
  4. [LinkedIn, October 2026] Aditya Nigam’s Post on AI architecture | https://lnkd.in/p/eAxbcgFF
  5. [Bloomberg, retrieved 2026] Anant Prakash, Synthomer PLC: Profile and Biography | https://www.bloomberg.com/profile/person/19532785
  6. [CB Insights, retrieved 2026] Funding round data | retrieved 2026
  7. [LinkedIn, September 2025] Raymond Burke’s Post on funding | https://www.linkedin.com/posts/r-burke_personal-post-im-passionately-supporting-activity-7371204067664072704-ENj8
  8. [Moving People, retrieved 2026] Article on Odysse revenue and Bolt deal | retrieved 2026

Read on Startuply.vc