Technical honesty
We state what is known, what is uncertain and what evidence would change the plan.
The firm works across domain adaptation, evaluation, agentic systems, retrieval and ML infrastructure, with direct senior involvement throughout each engagement.
Founder
Principal engineer
Avinash Mynampati
Founder & Principal AI Engineer
India → Global
Founder
Founder & Principal AI Engineer
Avinash is an AI and machine-learning engineer working across domain-specific language models, applied ML systems, training infrastructure and research prototyping. His professional experience includes roles at Juspay and Codingmart, alongside current work on specialised language-model systems for semiconductor engineering.
He founded Asthra AI Labs to work directly with teams tackling technically difficult problems—combining research judgement, engineering execution and meaningful evaluation from the earliest stage of a project.
Founder & Principal AI Engineer. Focused AI engineering and consulting across domain models, evaluation, agents and infrastructure.
Work on specialised language-model systems for semiconductor engineering, including adaptation, data and evaluation concerns.
Professional engineering experience in a fintech environment.
Professional experience across product engineering and applied software work.
Data Science specialisation, with international academic experience at EPITA, Paris.
Applied AI systems, research prototypes and production-oriented ML workflows.
Asthra operates as a focused, founder-led firm. Projects are accepted selectively and senior technical involvement remains direct.
Based in Andhra Pradesh, India. Working remotely with teams across regions and time zones.
We state what is known, what is uncertain and what evidence would change the plan.
Work is structured around evaluation, checkpoints and decisions—not open-ended activity.
Sensitive domain data and proprietary context are handled carefully and discussed only as needed.
You work with senior technical judgement rather than layers of account management.
We prefer systems and artefacts that remain useful after the engagement ends.
Short-term demos are not allowed to create long-term operational debt by default.
Next step
Bring the domain, constraints and current system. We will help determine what is feasible, what should be measured and what is worth building.
Initial conversations are exploratory and confidential.