Aurelia Ayala Usma Email & Phone Number

Bogotá, Bogota, Co

Bogota, D.C., Capital District, Colombia

Colombia

- Developed a bilingual Q&A chatbot app in Databricks to answer questions about procedures and functions of the Colombian Institute of Family Welfare (ICBF in Spanish). Web scraping and data pre-processing was carried out using Selenium, BeautifulSoup, Spark, Pandas, and Delta Lake. The chat model was built using LangChain implementing FAISS Vector Store and OpenAI APIs for text search and chat completion. The application backend is served via REST API hosted in Databricks. A Streamlit front-end was created to allow for the chat functionality.- Created a text classification workflow for news headers for an internal Sales Intelligence application. Three custom distilBERT-based classification models were fit and tuned using the HuggingFace Transformers library until a minimum precision of 0.8 was achieved. These models were served via FastAPI hosted in a microservice architecture.- Implemented and deployed an end-to-end financial credit risk prediction engine using Lending Club’s P2P lending dataset. The ETL process was performed using Pandas and imblearn, sourcing data from AWS S3. Five different prediction models were developed and tuned using Scikit-Learn, LightGBM, XGBoost, and Keras libraries. The models' explainability was achieved using the SHAP library. These models were served via FastAPI in an AWS Lambda environment, and a GUI in Streamlit was created to allow human interaction with the API.

Colombia

- Established and implemented Data Stewardship policies for Salesforce CRM with the data Governance Committee aiming to avoid introducing data gaps and inaccuracies in the creation of potential new customer records, and to solve existing data issues in existing records using BigQuery and Python.- Solved 1298 data quality issues in Salesforce accounts belonging to customers that bring >= 1M USD in revenue to the company.- Manually curated 659 Salesforce accounts of potential new customers.

Colombia

- Created an automated strategy for the data enrichment of >6K Salesforce account records using BigQuery, Python scripts, and Clearbit data provider to bridge data gaps found in the CRM database that hindered financial reporting and marketing efforts.- Provided ad-hoc analyses to Sales Operations team on customer data distributed across CRM, subscriptions, and Data Governance systems such as Salesforce, Zuora, and Reltio.

Colombia

- Lecturer for the "Data Analysis and Big Data: Analytics for Business Solutions" Diploma course- Topics covered: Python Programming, Neural Networks, Support Vector Machines, Databases, SQL, Big Data Environments.

Colombia

- Redesigned and maintained pricing algorithms for the voice and video products that drive ~90% of the revenue of the company. This involved data modeling, A/B testing, and back-end development in Python. This resulted in a 45% YoY increase in revenue for emerging non-English Voice projects and the launching of the Video product line in Q1 2021.- Formulated probabilistic models to forecast one-year revenue and one-year survival probability (churn) of the > 10K customers of the company, and automated them in a Python module that fed the analytics infrastructure. This enabled the CEO and Sales teams to plan ahead strategic decisions for customer acquisition and retention during 2021 and 2022.

Colombia

- Led a team of 3 people in the creation and maintenance of the entire Business Analytics infrastructure of the company. This included the entire data cycle from Data Capturing (Segment, Snowplow), ETL (AWS DMS Lambda), Warehousing (Redshift, AWS Postgres RDS), (K)PI standardization and logic development, to Visualization/Dashboarding (AWS EC2, Metabase). This work enabled the entire company to make reliable data-driven decisions with > 360 high-quality (K)PIs maintained by the team.- Built ad-hoc reports for Product, Marketing, Sales, and Customer Operations teams to support them in shaping the annual and quarterly strategy. This work mainly involved time-series statistical analysis and data treatment in SQL and Metabase.

Bogotá D.C. Area, Colombia

- Reconstructed the whole genome of two plant pathogens species of the genus Phytophthora using 16.7 GB of PacBio SMRT and Illumina whole-genome DNA sequencing datasets. This work resulted in two publications reporting the open-source genomic and computational resources produced for these two important agricultural pests. The project was done in collaboration with researchers from the CIRAD - Université de Montpellier and the Universidad Autónoma de Manizales.- Carried out two genome-wide association studies (GWAS) in US, Mexican, and Colombian natural populations of Phytophthora infestans to investigate the genetics of the resistance of the pathogen to mefenoxam fungicide, the thermal adaptation in climate change simulations, and the sexual reproduction determinants of the individuals. Generalized linear models (MLMs and ANOVA) were used for association between genetic variations and phenotypes. The project was done in collaboration with researchers from Cornell University, the University of California – Riverside, and the University of Hawai’i, Mānoa.- Affiliations: Research Group in Computational Biology and Microbial Ecology (BCEM) and Laboratory of Mycology and Plant Pathology (LAMFU).

Bogotá

- Scientific translator for Biomedical Industry. - Specialty areas: Pharmaceutical Documentation, Medical Devices, and Regulatory Affairs.

Bogotá, Colombia

- Performed anatomical identification of 4 tropical subfossil and modern woods from Colombia, and 10 modern woods from Chinchero Archaeological Center, Perú.- Analyzed δ13C isotopic data derived from a subfossil wood sample for paleoclimatic and biogeochemical inference.- Designed PCR primers targeting common chloroplast markers (trnL, matK, etc.) for molecular identification of modern and ancient woods.

Bogotá, Colombia

- Isolation of microorganisms resistant to pollutants in soils and wastewaters from oil-exploitation activities.- Perform functional experiments of biorremediation of petroleum pollutants and heavy metals upon the previously mentioned isolates.- Standardization and implementation of a spectrophotometric method for barium quantification.

Bogotá, Colombia

- Lecturer of Biology and Biochemistry for Health Sciences and Environmental Engineering students.- Curriculum evaluation and test design of Biology and Biochemistry courses.

Bogotá, Colombia

Molecular biology wet-lab contestant. Team Colombia at the International Genetically Engineered Machine Competition (iGEM 2012). http://2012.igem.org/Team:Colombia- Primer design, gene amplification, and molecular cloning of genes from Aliivibrio fischeri.- Synthetic DNA construct evaluation and design.

Bogotá, Colombia

Molecular biology wet-lab contestant. Team Colombia at the International Genetically Engineered Machine Competition (iGEM 2012). http://2012.igem.org/Team:Colombia- Gene amplification, and molecular cloning of genes from Aliivibrio fischeri.- Semi-quantitative RT-PCR gene induction assays.- Scanning Electronic Microscopy of leaf tissue colonized by bacterial cells.

Master’S Degree, Computational Biology

Graduate program with emphasis in Bioinformatics, Scientific Programming, and Data Analytics.

Data Science

Fellow of the Data Science for All (DS4A) / Colombia 6.0 Program

Bachelor Of Science (B.Sc.), Biology, General

Bachelor Of Science (B.Sc.), Microbiology, General