About Pharos Production

Who we are

Pharos Production is a software development and AI engineering company founded since 2013 by Dmytro Nasyrov (PhD AI, 23 years engineering experience, Kyiv Polytechnic Institute lecturer). Our team ships production systems for founders and enterprises who need real engineering depth, not generic agency work.

We work mostly with FinTech CTOs, AI product leaders and enterprise architects who need real engineering depth - not generic agency work.

We decline roughly 30% of incoming RFPs when we believe AI or a given technology is not the right answer. We prefer to recommend the honest path rather than sell more scope. That posture is why our rating is 5.0 across 315+ verified reviews^[1] on Clutch, GoodFirms, Google, Glassdoor and G2 and why our average client engagement runs more than three years.

What we do

• AI engineering: LLM integration, agents, RAG systems, ML platforms, AI infrastructure
• FinTech development: payment systems, banking platforms, compliance tooling (see Pleenk - Secure Payments Platform)
• Blockchain and Web3: protocol engineering, DeFi, NFT, crypto wallets (see NoMoreBets, GridTradeX)
• Custom software: enterprise platforms, SaaS, marketplaces, mobile apps (see Gambit Stream)
• Staff augmentation and CTO-as-a-service

By the numbers

• Founded 2013 in Kyiv, Ukraine. Primary sales office in Las Vegas, Nevada
• 90+ engineers, designers, product managers and compliance specialists
• 70+ apps delivered to 200+ clients across 18 industries
• 315+ aggregated client reviews^[1] rated 5.0 across Clutch, GoodFirms, G2 and Glassdoor
• Aligned with ISO 27001:2022^[2], SOC 2 Type II^[3], HIPAA-aligned, PCI DSS-aligned and GDPR frameworks

Our approach

We apply The Pharos Engineering Loop - our named delivery methodology: weekly client demos, written decision records, reference architectures and post-mortem culture. Every project has a named CTO-level lead, an assigned product manager and a documented exit plan. Clients own their source code, infrastructure credentials and roadmap artifacts from day one.

What we have learned

After 13 years and 70+ delivered apps, the lessons that shaped Pharos are uncomfortable: most "AI projects" fail because the data is dirty, not because the model is weak. Most blockchain projects fail because token incentives outpace product fundamentals - we learned this building Kimlic's blockchain KYC protocol. Most enterprise integrations fail because requirements freeze before user research - Nextcheck's KYC platform is one example. We decline 30% of incoming RFPs not to maximize margin, but because we have learned that scoping the wrong project burns both client capital and engineering velocity. Our post-mortem culture exists to make sure we keep learning these lessons in public, not in private.

"Most AI projects fail because the data is dirty, not because the model is weak."

Dr. Dmytro Nasyrov Founder & CTO, Pharos Production

Selected publications

Dmytro publishes engineering analysis on AI economics, drift detection and FinTech compliance:

• All articles on Medium

Ready to ship? Talk to engineering→

Last reviewed: June 6, 2026