I am Jonathan Soto, Vice President of Engineering in the company Visage.jobs You will find below an internship offer in the field of language processing. The position is based in Nantes
Do not hesitate to contact us on talent at visage.jobs if you are interested
Regards Jonathan soto P.S The internship offer: Visage
Visage is an intelligent talent platform that helps companies recruit more efficiently. Visage sourcing technology blends 4,000 Sourcers and AI to find talent within hours. We are revolutionizing the way recruitment is done at companies like L'Oreal, Siemens, Orange, etc.
Our Venture-backed startup is headquartered in San Francisco.
You will join a team that focuses on building an amazing product *coming straight out of Silicon Valley*! We are a 20 people tech team with many skills : data, devops, dev, q&a Smart Email Assistant - Project Motivation
When clients review profiles sourced by our community, they can either disqualify them, contact them, or save it for later. Our platform allows users to contact and follow-up their prospects easily with an *email template auto-filled* with job title and its location, candidate name, and current position.
However, we prompt users to *personalise these cold emails* to increase open and response rates. They should make sure they are interesting, appealing, and written in a tone that *reflects their company’s culture and values*. The more interesting, the higher the chances for a recipient to become a potential candidate. Last but not least, the subject line is essential. It doesn’t matter how good an email is if it’s never opened.
Other strategies can be impactful to *grab passive candidate’s attention* such as considering the recipient’s time zone. Your Role *Post-internship PhD*
Our ultimate goal is to *develop a personalized cold email generator*. The content of the email has to meet general guidelines in terms of length, clarity, and information provided. But besides what we say, the way we say it is what *makes the difference*. The tone of the voice must reflect the user brand culture in order to *attract the right audience*. For example, a traditional company will go for an opening such as “Dear Mr. Smith”, while a disruptive start-up looking for young graduates would prefer “Hey John”.
This new feature would be a clear *competitive advantage* for Visage. It will help our users to both save time and increase the engagement rate. This is a *long-term and ambitious project* that will be subject to a *PhD*. *Internship Mission*
This internship will set the basis for the thesis in terms of data, exploratory analysis, modeling, and research framework. You will leverage *5 years of data* including *thousands of email exchanges*. Under the supervision of our senior data scientist, your mission will consist of several stages :
- Data Gathering :
- Engagement rates (target): % open, reply and lead rates.
- Email-related data: title, body, signature, length, sender email
type, user, and recipient time zones...
- Use Data Science techniques to tag :
- *Level of Personalization*: Determine which emails were
personalized by the user in historical data. Emails can be compared to
templates provided by Visage. Text similarities will be evaluated using
either naive methods, embeddings associated with classical Machine Learning
models or Deep Learning.
- *Spammy email content*: Investigate the relationship between
bounced emails and their content. Were some messages rejected due to some
- Both subjects may lead to *model prototypes and APIs creation*.
- Exploratory analysis and data visualization :
- Investigate the *relationship between emails’ attributes and
- Design and *develop visualizations* to provide users metrics to
track the impact of their personalization effort.
- Start some literature search :
- Start exploring what methodologies and models other researchers
have already developed in the field of *text generation*.
- Investigate the capability of the natural language models such as
*GPT-3* or other promising models to generate short emails in
different writing styles.
Your Profile and Skills
You *think before you code*, you like to have passionate discussions about the *best ways to solve a problem*. You understand that technology is not an objective but a way to get the most efficient solution to a problem.
You are a student in your *last year of engineering school* and you are *seeking a PhD* after graduation. You have knowledge and experience in:
- *Machine Learning*
- *Neural Networks*
- *Natural Language Processing*
- *Database* Systems
- *Python* Programming and main libraries (Pandas, Numpy, Scikit-learn,
We hire and keep people who are constantly overachieving. However, we don’t hire jerks, as brilliant as they can be. You will be part of a pro sports team that knows that excellence comes from being open-minded and always challenging ourselves. Our values are :
- *Visionary*. We anticipate where the world is going and we build
- *Meaning*. We don’t sacrifice our long term vision and beliefs for
short term returns. We work the hardest when nobody is watching.
- *Honesty*. We practice radical transparency: direct and honest in
communication and sharing of company strategies.
- *Balance*. Whilst we work intensely, we value life balance. We take
time to rest and refresh.
- *Impact*. We cover a lot of ground on truly important work. We focus
on great results.
talent at visage.jobs -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/html Size: 6809 bytes Desc: not available URL: <https://mailman.uib.no/public/corpora/attachments/20220102/282093fe/attachment.txt>