Perspicace |pɛʀs.pi.kas|
Perspicace Inc. is a management and technology consulting, software and services company operating in the fields of artificial intelligence (AI) and high performance computing (HPC). Perspicace provides end-to-end Contextual AI design services, working with you to develop your AI-driven business strategy, mitigate business and technical risks while building capacity, and drive innovation through accelerated time-to-results.
Our mission is to advance the use of AI for positive, ethical results in the industry, develop societal capacity through expertise transfer, and realize material benefit to our people, our community, and our environment.
Photo by Pixabay
Who We Are
We represent 30+ years experience developing advanced analytics solutions which meet real customer needs. We are human experts dedicated to the ethical use and development of AI.
We believe in open source, open models, open platforms, and open research. We publish our models as containers so the community can benefit, contribute and continue to innovate.
We work with the community to improve reproducibility and transparency, improving methods, accuracy and performance of analytics pipelines.
We bring context, mindfulness, transparency and accountability to mitigate bias and confront hate. We incorporate disparate data, algorithms and viewpoints to deliver rich, contextual AI.
We continually reinvest in our expertise, closely monitor customer requirements, and push the envelope to improve quality, cost and time-to-results.
We carefully curate our industry partners amongst a select group of like-minded institutions with similar values and demonstrated ethical behavior.
Fundamentally, we are empiricists – we apply the scientific method, identify emerging patterns, and use real-world results and metrics to constantly challenge and improve.
Kristina Kermanshahche
Founder & CEO
Kristina Kermanshahche is an entrepreneurial executive and industry thought leader, who combines deep vertical domain knowledge with strategic business acumen and strong technical expertise to deliver transformative solutions to complex needs. Working with customers at the leading edge of adoption, she identifies key inflection points and industry trends years in advance; developing industry partnerships to overcome barriers to adoption fuels market expansion worldwide.
During her twelve-year tenure as Intel Corporation's Chief Architect, Health & Life Sciences, Kristina set the vision, assembled experts and technically led cross-geo AI investments in such diverse areas as image analysis (using Caffe & TensorFlow MCNNs), cryo-electron microscopy, neurodegenerative imaging, genomics analysis, predictive clinical analytics, and applied machine learning to improve manufacturing processes. Fostering strong collaborations with commercial and open source teams, Kristina consistently enabled scientific breakthroughs worldwide by delivering order of magnitude gains in performance and scalability.
Kristina developed numerous HPC AI on-prem and cloud-based cluster reference architectures, driving significant results in workload characterization, code modernization, benchmarks and infrastructure sizing. She also co-developed a healthcare semantic interoperability engine and microservices functionality to address the key challenges of data injest and preprocessing (US & Int’l Patents US11858842). In addition, Kristina served as trusted advisor to governments and institutions worldwide on national healthcare and precision medicine initiatives, HPC AI, regional secure cloud computing, and shared services strategy.
Prior to joining Intel, she developed industry leading natural language processing, speech recognition, and high availability analytics services (99.999% uptime), spanning cloud service providers, financial services, manufacturing, commercial, and transportation. Throughout her career, Kristina has developed commercial products & services, working across the spectrum of Fortune 50 and strategic startups worldwide.
Ms. Kermanshahche is a frequent speaker on AI HPC Emerging Patterns & Best Practices, and High Throughput|Low Latency Analytics. Her research interests include grid cell navigation and the potential for composable models (generative, reinforcement, meta) to produce self-learning systems.
For more information on Kristina’s industry results, please explore her LinkedIn profile.
