Technical Advisory Services
Expert guidance for early-stage deep tech and AI startups, from first prototype to a funded, compliant product.
Where I focus
The work early-stage teams and their investors come to me for most. Start here.
Not sure which fits? Book a free 15-min intro call
Work behind the advice
The advisory comes from shipping in regulated, physical domains, not from a slide deck.
Star Hopper Studio
AI operations & growth advisory for an award-winning film production studio
Founders offloading busywork to AI systems they own and run themselves
Bearcover
Autonomous patrol robots for care facilities
CE marking & ISO 13482: certified and deployed in German nursing homes
CarbonFreed
AI for grid certification of PV & wind systems
Compliance automation in one of Germany's most regulated energy processes
VoltVogel
Co-founded EV-charging robotics
Hardware from first prototype to paid pilots, and an honest wind-down
“Abhi's helping us use AI to offload the busywork so we can focus on the creative side, and to think about growing the studio like a startup. Nothing is a black box — we run everything ourselves.”
Frequently Asked Questions
What types of startups do you work with?
I work with pre-seed to Series A deep tech and AI startups. This includes companies building agentic AI systems, robotics, hardware, SaaS, and regulated tech. I bring 8+ years of experience across hardware, software, and AI.
How long does an advisory engagement typically take?
Project-based engagements range from 2-5 days (pitch deck review) to 4-6 weeks (system design or PMF sprint). Retainers are ongoing monthly arrangements. Most feasibility assessments take 2-3 weeks.
Do you offer ongoing support?
Yes. I offer monthly retainers for Fractional Engineering Lead, Compliance Advisory, and Pre-PMF Technical Advisory. These provide ongoing strategic guidance, architecture reviews, and technical support as you scale.
Can you help with EU AI Act compliance?
Yes. My compliance background covers CE marking and ISO 13482 (Bearcover, robotics), grid certification for PV/wind systems (CarbonFreed), and regulated product development (VoltVogel, EV charging). I also wrote The EU AI Act Field Guide, so I can help with risk classification, gap analysis, and building a compliance roadmap — and I'll be direct about what requires a notified body versus what you can self-assess.
What's included in technical due diligence?
Code review, architecture assessment, technology stack evaluation, security and compliance audit, scalability analysis, team capability evaluation, and a detailed report with risk ratings and recommendations.
How do you ensure confidentiality?
NDA available. I work with sensitive technical information regularly and follow GDPR best practices. For code review, we can use secure access methods and I never retain copies of your code.
Do you work remotely?
Yes. I'm Berlin-based but work with startups globally (EU and US). All advisory work can be done remotely via video calls, async communication, and shared documentation.
What if we're not sure which service we need?
Book a free 15-minute intro call. I'll help you identify the right engagement—whether it's a feasibility assessment, due diligence prep, architecture design, or ongoing advisory.
Do you offer equity-based arrangements for early-stage startups?
Yes. For pre-seed startups with limited capital, I can structure engagements with an equity component. This aligns our incentives—I'm invested in your long-term success. Equity discussions are handled respectfully and transparently. Reach out to discuss if this might work for your stage.
Can you do hands-on development as part of advisory?
Yes. Advisory engagements can include hands-on development and design when needed—prototypes, architecture implementation, code reviews, or targeted build-outs. I'll scope what makes sense for your engagement.
Related Tools & Resources
Quick Links
Related Articles
The Model Is a Component: Why AI Launches Stopped Mattering for My Stack →
AI/ML · 8 min read
Four frontier AI models shipped in seventeen days this July, and my stack needed one config change. ...
The VLM That Scored a Collapsing Robot 62/100 →
AI/ML · 16 min read
After metrics kept missing degenerate gaits and LLM-iterated rewards hit the survival cliff, we trie...
When LLMs Iterate on Rewards: A Negative Result From Humanoid Locomotion →
AI/ML · 15 min read
We ran NVIDIA's Eureka methodology — an LLM iteratively proposing reward functions for an RL agent —...
Interested in working together? Get in touch or schedule a consultation.