CRA-W Career Mentoring Workshop“Getting Started in the Lab: Tips for Surviving

the First Two Years” (L-track)

June 4-5, 2011, San Jose, California

A.J. Brush, Microsoft ResearchEvi Dube, Lawrence Livermore National Lab

Raquel Romano, Google

You bring a solid foundation to build skills and experience

• Utilize this foundation as you begin to build the layers of your career

University Connections

Post Doc Experience

IndustryCollaborations

New/Fresh Perspective

Think in terms of building blocks for success

Get a mentor

Learn about your environment

Network

Deliver

Understand Stakeholders

Give Talks

Publish

DeliverDeliver

Prioritize based on the long term view

• Use the building blocks as guides• Deliver high quality work • You are the Champion of your work

Largest % time …..... Smallest % time

A.J. Brush

Software Developer at Tripos1998 - 1999

Senior Researcher, MicrosoftHuman-Computer Interaction (HCI)2004 - present

B.A. in Computer Science from Williams1992- 1996

Ph.D. in computer science from UW1996 – 1998, 1999 – 2002

PostDoc, UrbanSim project at UW2002 - 2004

I thrive on finding ways for computers to help people solve everyday problems

Microsoft Research

Established in 1991More than 850 researchers, 55 areas of research

To advance the state of the art in computing through a combination of basic and applied research.

MSR is a small part of Microsoft ( < 1% of employees)

A.J. Brush

How do I spend my time?

We are evaluated on:• Research impact• Product impact• Patents

Changes week to week

Things I do:• Research Projects• Go to meetings/answer email• Consult to product groups• Service (Internal/External)

A.J. Brush

How do I spend my time?

5040

10

Last Summer

ProductResearchService

10

4050

Last Week

20

70

10

This Summer

A.J. Brush

I develop tools for people to extract useful information from massive amounts of data

BA, Math, Harvard, 1992

MS/PhD, MIT, 1995/2002

Postdoc, Computing Research Divison,

2004-2007

Software Engineer, OCR/Crisis Response

(2007-present)

Raquel Romano

Google & Google ResearchRaquel Romano

How do I spend my time?

• LBNL: investigate application of machine learning to scientific data analysis

• Google OCR Team: research, build, & train scalable text detection and multi-language recognition systems

• Google Crisis Response: find accurate, relevant information from many data sources during emergency events

Raquel Romano

I juggle several diverse projects in two key roles

Technical LeadershipProject Management

Energy Prog. Models HPCSISCR

Deputy

Ops Manager Evi Dube

LLNL Researchers use HPC to answer difficult problems

High performance computing accelerates innovation

American Competitiveness• Energy• Manufacturing• Science

Evi Dube

LIFE AT LABS

Life at a Government Lab

• Basic or applied research• Soft money, block grants, budgeted

funds• Flexibility: can set your own hours• Managed environment

May be harder to develop your own research program vs. working on an existing program

• Taxpayer money: limit on daily perks!

Applied Research

• Team projects Junior researchers are generally members of a

team Team will most likely have some

goals/deliverables that are not exclusively research

The research will frequently be a team effort as well

• Setting research agenda Usually requires some time at lab Must be relevant to lab’s strategic mission

Industry Labs

• Wide range of opportunities PARC, Microsoft, IBM, AT&T, Nokia, Motorola,

Google, Amazon, Intel ….• Dimensions they differ

Research flexibility: Do you choose your own projects or get direction from product groups?

Funding models (e.g. separate division, sponsored by product teams)

Participation in research community (e.g. publishing)

Team/Research group structure

Pro’s/Risks

• Pro’s: Funding “taken” care of Typically well-resourced (travel, etc.) Ability to have direct impact on

products/people Relatively easy to adjust research direction/try

new areas• Risks:

Labs can change (e.g. Intel Research labs closed spring 2011)

Companies sensitive to economic climate

STEPS TO SUCCESS

Starting Out: Mentors

• Find mentors You may or may not have a formal mentor Different mentors for different activities

(research, program activities, lab politics, etc.). Include someone outside your reporting chain! “1/2 hour of your time” Ask for advice, tips, introductions, stories.

• Participate in research community Labs can be insular Go to conferences, give talks

Starting Out: Visibility

• Working in many different areas can have benefits But do not become so fragmented you can’t do

your best on each task.• Establish a reputation at your lab. Be visible.• Establish your area of expertise and find your

community.• Find what conferences you want to publish in.• Community service (program committees,

reviewing) are not rewarded as much as in academia, but necessary for your growth as researcher.

Starting Out: Publish

• “Publish or perish” is not just for academic researchers

• Some non-academic institutions may not put much emphasis on publishing, but research community as a whole does.

• Career options are severely limited if publications stop.

Getting Known Inside the Lab• Produce great work and make it known

Write papers/technical reports Give talks within the lab. If your lab has an education or

outreach office, get to know them. Your manager(s) should be praising you to others. Make

it easy for them by providing updates, slides, demos. Share appropriate credit with your collaborators. Seek collaborators. Start reading groups and invite colleagues. May find

future collaborators.• External recognition will usually come before

internal recognition Make sure management hears about it!

Getting Known Outside the Lab

• Write workshop papers and posters, in addition to conference and journal articles.

• Talk tours Self-invitation (“I’ll be in the area”)

• Proposal review panels, journal refereeing, conference program committees: volunteer yourself.

• Invite others to visit and give talks.

How do you decide what to work on?

• You may be assigned to a project initially• If you’re free to choose, think about your

criteria for choosing a project, e.g., one best aligned with your research interests? one that will teach you the most? one best aligned with the lab mission? one led by a PI you want to work under? one with a team you would enjoy working with?

• Initial project: you are on probation, so make a good impression

Research Independence

• When is the right time to take the initiative, to launch new projects, or to take on a leadership role? Senior collaborators see opportunities

and give you leads Acquire your own funding Visibility and accomplishments give you

the necessary credibility

LESSONS FROM OUR EXPERIENCE

Things I’m glad I did

• Internally: Worked on projects with different people Learned the culture Got mentors for specific situations Organized the MSR Women’s group with Lucy

Vanderwende, intern talk series • Externally:

Publishing/Workshops/Visibility/Networking “Volunteer” for Program Committees SIGCHI VP for Membership Picked some smaller conference to “focus on”

A.J. Brush

Things I wish I had done

• Meet and greet with lots of people right after I started

• A little bit more coherence/plan with projects around clear theme

• Valued my time, weighed opportunities, killed things sooner

• Figured out sooner when to say “No” (I’m getting better at this )

• Aggressively avoid meetings

A.J. Brush

Things I’m glad I did

• Pursued my intellectual and personal passions, both during and after graduate school.

• Changed course when the time was right in spite of uncertainty.

• Used each new experience as a learning opportunity for the next.

• Found a way to network without “networking” via social groups and volunteer activities at work

Raquel Romano

Things I wish I had done

• Shared work in progress before it was complete or polished, e.g. at team weeklies or with people outside my immediate group.

• Joined team projects rather than working too independently.

• Figured out when to shift gears on a project earlier rather than later.

• Found and used more mentors.

Raquel Romano

Things I’m glad I did

• Took risks to go into new areas not typical for computer scientists

• Delivered in everything asked to do• Active in assignments• Spent some time in Washington D.C.

on assignment”• Pursued the Ph.D. and the PMP

Evi Dube

Things I wish I had done

• Understood the value of a mentor for myself earlier in my career

• Learned to listen to others with a more open mind earlier in my career

• Aggressively avoid meetings

Evi Dube