Firm Description
Mix is a premier AI companies supplier, dedicated to co-creating significant impression for its purchasers by means of the facility of information science, AI, know-how, and other people. With a mission to gasoline daring visions, Mix tackles important challenges by seamlessly aligning human experience with synthetic intelligence. The corporate is devoted to unlocking worth and fostering innovation for its purchasers by harnessing world-class individuals and data-driven technique. We consider that the facility of individuals and AI can have a significant impression in your world, creating extra fulfilling work and initiatives for our individuals and purchasers. For extra data, go to www.blend360.com
Job Description
We’re on the lookout for a Knowledge Scientist to drive efficiency optimization for our purchasers AI Engine. This high-impact position will deal with complicated computational bottlenecks and contribute to the re-engineering of simulation, coaching, and post-processing pipelines for large-scale industrial AI fashions. The candidate will collaborate with a multidisciplinary staff to ship quantifiable enhancements in velocity, scalability, and reminiscence effectivity.
Duties:
-
Analyze and optimize AI engine code, specializing in eradicating efficiency bottlenecks in simulation, coaching, and post-processing workflows.
-
Refactor sequential code and nested loops into environment friendly, vectorized operations, leveraging superior information in linear algebra and matrix decomposition.
-
Diagnose and resolve computational inefficiencies associated to GPU/TPU, non-vectorized knowledge dealing with, and combined framework operations (JAX, NumPy, Pandas).
-
Develop and implement options for ODE (strange differential equation) solvers, optimization algorithms, and batch processing methods.
-
Lead root trigger evaluation for efficiency limitations and suggest different algorithmic methods (together with MILP/LP, decomposition methods, and different ODE solvers).
-
Information remediation and refactoring efforts, together with reminiscence optimization, JIT compilation, and knowledge kind standardization.
-
Doc enhancements, monitor ongoing efficiency, and contribute to a roadmap for additional scalability enhancements.
{Qualifications}
-
Masters in Arithmetic, Operations Analysis, Laptop Science, or associated quantitative subject.
-
Deep experience in matrix decomposition, numerical optimization, and ODEs, with a demonstrated skill to use these in real-world computation.
-
Sturdy proficiency in Python, with hands-on expertise in JAX, and expertise with GPU/TPU acceleration.
-
Familiarity with ensemble strategies, batch and vectorized computation, and reminiscence administration in massive datasets.
Most well-liked:
-
Data of MILP, LP decomposition, and different ODE solvers (e.g., diffrax, odeint).
-
Expertise with CI/CD practices for ML/AI pipelines, profiling instruments (e.g., cProfile, memory_profiler, jax.profiler), and efficiency benchmarking.
Extra Data